Digital multi-user radio telephone system - has master station radio linked to sun stations which are time division multiplexed onto single channel

(19) DENMARK (11)

<img file="DK175353B1_D0001.tif" />

(12) PATENT WRITING

Patent and

Trademark Office (51) Int.CI⁷.: H 04 B 7/26 H04J 3/00 (21) Patent Application No: PA 2002 00209 (22) Filing Date: 2002-02-12 (24) Running Date: 1985-09-20 (41) Aim. available: 2002-02-12 (45) Patent Notice bkg. on: 2004-09-06 (30) Priority: 1985-03-20 US 713925 (73) Patent Owner: InterDigital Technology Corporation, 913 Market Street, Suite 802, Wilmington, Delaware 19801, USA (72) Inventor: Erich Paneth, 4178 Decord Street No. 68, San Diego, California 92122, USA Mark J. Handzel, 6750 Breadnell Way no. 39, San Diego, California 92117, United States (74) Plenipotentiary: International Patent Bureau A / S, Hoge Taastrup Boulevard 23, 2630 Taastrup, Denmark (54) Title: Digital Wireless System (57) Summary:

Wireless digital system comprising a base station (11) in communication with telephone lines and several subscriber stations (10) for simultaneous transmission of information signals between the base station and each of the several subscriber stations (10) via radio frequency channels (RF). The system includes converting means (15) at the base station (11) for connecting the relevant telephone lines and converting the analog information signals received from the telephone lines into digital signal samples and for converting digital signals received from the subscriber stations (10) to analog signals for transmission to phone lines. Further, the system comprises compression means (17) connected to the converting means (15) for simultaneously compressing the digital signal samples arriving from the converting means (15) to provide separate compressed signals, channel control means (18) connected to the signal compression means (17) to sequentially combine the separate compressed signals from the signal compression means (17) to a single transmit bit stream, wherein each of the respective compressed signals occupies a repetitive, sequential position in the transmit bit stream, transmitting and receiving means at both the base station (11) and at the subscriber stations (10) continue to provide direct communication between the base station (11) and the subscriber stations (10) via the radio frequency channels each subscriber station (10) operates in half-duplex mode within a raster by time multiplexed access, transmitting in one part of the raster and receiving in another part of the raster, and means for providing a periodic exchange of control information between the subscriber station and the base station with which it communicates forward and return signals during this communication and at the same time intervals as the forward and return signals, which control information includes the instantaneous status of the connection between the base station and the subscriber station, the quality of the connection between them, power and timing, and on the basis of that, provide adjustment signals from the base station to the subscriber station with which it communicates, at the same time intervals as the forward and return information signals.

<img file="DK175353B1_D0002.tif" />

<img file="DK175353B1_D0003.tif" />

(UHF RADIO)

GROUND STATION • 10

<img file="DK175353B1_D0004.tif" />

IN-^weij PHONE- I * COMPANY! CENTRAL ii

The invention relates to a digital telephone system for simultaneous, wireless transmission of information signals over at least one radio frequency channel between a base station and a plurality of subscriber stations, and wherein the base station simultaneously receives such information signals over telephone lines and simultaneously transmits these signals over the radio frequency channel, and wherein the base station and the subscriber stations each have one. transmitter / receiver.

These subscriber stations may be stationary or mobile. The base station is connected to an external information network. The information signals may be voice, data, facsimile, video, computer or instrument measurement signals. The mobile stations may be in relatively fast or relatively slow motion.

The publication NEC Research and Development, no.

76, pp. 24-35, published Jan. 1985, discloses a system based on Time Division Multiplex Access (TDMA) technology, which is intended to replace conventional analog telephone systems. An important feature of this known system is that it uses the Digital Radio Concentrator System (DRCS) technique, which allows operation between a relatively large number of cells and the base station.

One disadvantage of such a system is primarily to be seen in that it serves subscribers through repeater stations, that it uses ordinary voice signal coding, and that therefore only a relatively limited number of subscribers can be operated at a time.

The invention is directed to a digital telephone system which does not exhibit these disadvantages and for this purpose the system according to the invention is designed as defined in the characterizing part of claim 1.

The system ensures spatial dispersion (diversity) between several cells at appropriate distances from each other, whereby relatively strong signals can be received, despite fading.

The base station operates over a number of radio frequency channel pairs. Each channel pair operation is implemented by combining a broadcast channel circuit for processing a given number of information signals received simultaneously over telephone lines, for simultaneous transmission to different subscriber stations over a given radio frequency channel, with a receiving channel circuit for processing a number of signals received simultaneously over a given radio frequency channel from 10 different subscriber stations, for the purpose of generating information signals for transmission over the connecting lines.

To each of the connection lines, separate converters are connected to convert the information signals received over the connection lines into digital signal samples.

The transmit channel circuit comprises a given number of separate signal compression circuits for simultaneous compression of the digital signal samples arriving from the different converters for producing a corresponding number of separate compressed signals, a channel controller coupled to the compression circuits for sequential combining of the compressed signals into a single bit stream . the respective compressed signals repetitively and sequentially occupying their respective interval positions in the bitstream corresponding to the associated compression circuit, as well as a unit for producing a transmit channel signal for transmission over a given radio frequency _v channel associated with the bit30 stream.

A central switches the respective converters to particular compression circuits among said separate compression circuits.

To the connection lines is connected a remote connection processor which responds to a call signal arriving over one of the connection lines to produce an interval allocation signal indicating which one of the separate compression circuits the switchboard must connect to one of the separate converters connected to said connection. connection line, and thereby assigns this line the range of bitstream associated with one of the separate compression circuits, which is thus connected through the central. This processor maintains a memory at which intervals are allocated in this way, and it interrogates the memory when a call request arrives, upon which it produces the interval allocation signal which provides for connection to a compression circuit not already assigned to another connection line.

To the remote connection processor, there is coupled a call processor which responds to the interval assignment signal to cause the control panel to complete the connection specified by the interval assignment signal.

The receiving channel circuit comprises a receiving unit for receiving an oncoming channel signal and for processing this signal to produce a bit stream of separate compressed signals in different repetitive sequential interval positions, a plurality of separate signal syntheses each associated with its interval position in the received bit stream. for the purpose of reconstructing digital signal samples from the separate, compressed signals in the associated interval positions in the received bitstream, and a channel controller for separating the separate compressed signals from the bitstream and routing the separated signals to the synthetic circuits associated with their respective time intervals from which the signals are separated.

Separate converters are connected separately to each other to convert digital signal samples to information signals transmitted over the respective connection lines. Each converter is connected to each converter through a common connection line.

The exchange connects the respective converters to specific synthesis circuits among said synthesis circuits.

The remote connection processor responds to the request signal received over that connection line5 and generates an interval allocation signal indicating which one of the synthesis circuits the switchboard must connect to one of the separate reconverters connected to said connection line, thereby allocating this interval in that line. the bit10 current connected to one of the synthesis circuits thus connected through the switchboard. The remote connection processor maintains a store in which the intervals are allocated in this way, and it interrogates the store when a call request arrives and then delivers the interval allocation signal to the call processor to provide connection to a synthesis circuit which is not already in one of the intervals another connection line.

The system of the invention uses advanced

LSI technology to achieve low cost, reliable high quality communication in the various applications. A preferred embodiment uses a fixed, centrally located base station to communicate with a large number of subscriber stations in a nearby geographical area. The central base station may be connected to the switchboard in a public telephone network through a private subscriber center switch (PBX) connected to arriving telephone lines. The subscriber stations in this system may be fixed, removable or mobile stations which can operate while in slow or fast motion. The subscriber stations communicate with the base station over UHF channels and with the users over conventional two-wire DTMF telephone equipment or via RS-232C or via non-conventional, e.g. four-wire tele35 stations. The system can replace existing wired local subscriber lines or be used to create quality telephone service in areas where fixed connections cannot be established or are not considered economical.

A characteristic of the system according to the invention 5 is the possibility of using time multiplexed approach (TDMA) as well as digital coding of speech signal in order to be able to use several frequencies simultaneously within a given network. On a given frequency channel (channel spacing 25 kHz), any number of high-quality voice circuits can operate simultaneously. To illustrate this, four such circles are used. This provides an advantage - in terms of spectrum and economy - over existing analog systems that, over a given frequency channel, only allow one communication at a time.

The measures that lead to the possibility of operating with fixed, mobile or portable stations at low cost are partly the use of digital low-speed coding of speech, with less than 16 kbit per minute. secondly, spectrally effective digital modulation techniques.

For example, the combination of a speech coding of 14.6 kbps with 16-level binary phase shift modulation allows for four simultaneous full-duplex calls on a single pair of 20 kHz wide channels with 25 kHz channel spacing, across the spectrum, especially within the bands

400-500 MHz and 800-950 MHz. Such a combination enables good quality conversations over distances of at least 20 km.

In order to compete with wire telephony, a much larger number of subscribers must be served, 30 than what a single pair of 25 kHz channels can carry simultaneously. For example, a system with 12 channel pairs and 47 simultaneous calls could serve 500 subscribers (ongoing calls plus pipes added - so-called off-hook plus onhook - the maximum contingent on the desired probability of peak blocking 35). Therefore, managing call requests from subscribers to obtain reasonable waiting times for connection establishment is also an essential feature of the invention.

The invention also relates to a wireless digital system comprising a base station in communication connection with telephone lines and several subscriber stations for simultaneous transmission of information signals between the base station and each of the several subscriber stations via radio frequency channels (RF), the system including converters at the base station for connecting the telephone lines concerned. and converting the analog information signals received from the telephone lines, for digital signal samples and for the conversion of digital signals received from the subscriber stations (10) to analog signals for transmission to the telephone lines, which system is characterized in that it further comprises compression means connected to the converter means for simultaneously compression of the digital signal samples arriving from the converter means. separate compressed signals, channel controllers connected to the signal compression means to sequentially combine the separate compressed signals from the signal compression means into a single transmit bit stream, each of the respective compressed signals occupying a repetitive, sequential position in the transmit bit stream, transmitting and receiving means at the base station as well as at the subscriber stations communication between the base station and the subscriber stations via the radio frequency channels, each subscriber station operates in half duplex mode within a raster by time multiplexed access, transmitting in one part of the raster and receiving in another part of the raster, and means for providing a periodic exchange of control information between the subscriber station and the base station it communicates forward and return signals with, during this communication and at the same time intervals as the forward and return signals, which control information includes the instantaneous status of the connection between the base station and the subscriber station, the quality of the connection between them, power and timing, and on the basis thereof provides adjustment signals from the base station to the subscriber station with which it communicates, at the same time intervals as advance s and the return information signals.

The invention is explained in more detail below with reference to the schematic drawing, in which fig. 1 is a block diagram of a radio frequency telephone system in accordance with the invention; FIG. 2 is a block diagram of an embodiment 20 of the base stations of the embodiment of FIG. 1; FIG. 3 is a block diagram of a subscriber station in the embodiment of FIG. 1; FIG. 4 illustrates the sequence of signals generated by the subscriber stations and the base stations to establish a connection between two subscriber stations; 5 various data processing modules implemented in the remote connection processor unit in the embodiment of FIG. 2; FIG. 6 shows the processing of incoming and outgoing messages in the baseband control channel of the remote connection processor in the FIG. 2; FIG. 7 processing the incoming and outgoing private subscriber exchange messages of the remote connection processor in the embodiment of FIG. 2; FIG. 8 shows the processing of the logger messages of the remote connection processor in the FIG. 2; FIG. 9 shows a memory card for the remote connection processor in the FIG. 2; FIG. 10 processing messages relating to

In the state of the radio control channel by means of the device shown in FIG. 5: message processing module shown; FIG. 11 the processing of messages relating to the state of the channel by means of the embodiment of FIG. 5, the message processing module shown in FIG. 12 is a block diagram of the subscriber interface unit of FIG. 3; FIG. 13 shows the signal interface between the private subscriber15 central and the speech encoder / decoder unit in the FIG. 2; FIG. 14 the signal interface between the subscriber interface unit and the speech encoder / decoder unit in the embodiment of FIG. 3; FIG. 15 shows the time relationships between the 13

PBX-VCU interface signals and those of FIG. 14 shows the STU-VCU signals, FIG. 16 shows the signal interface between the speech encoder / decoder unit and the channel controller both in the embodiment shown in FIG. 2 vi25 th base station and in the one in fig. 3; FIG. 17 shows the time relationships between the transmit channel signals in the FIG. 16 shows the VCU-CCU interface; FIG. 18 shows the timing conditions for the receive channel signals in the embodiment shown in FIG. 16 shows the VCU-CCU interface; FIG. 19A and 19B, respectively, the time ratios of transmit number block and receive speech block between VCU and CCU by phase shift modulation at 16 levels; 20A input and output data timing and content of the receiving channel between VCU and PBX (or STU) by 16-level phase shift modulation; 20B input and output data timing and content for the transmit channel between VCU and PBX (or STU) at phaseDK 175353 B1 16-level shift modulation; 21 is a block diagram of the channel controller both in the embodiment of FIG. 2 the base station shown in FIG. 3; FIG. 22 shows the functional, software-implemented design of the embodiment shown in FIG. 21, FIG. 23 is a timing diagram for transmitting the radio control channel data and the 16-phase phase shift modulated voice data over the transmitter for the embodiment of FIG. 22, FIG. 24 is a timing diagram for transmitting the radio control channel data and the 16-phase phase shift modulated speech data over the receiving bus for the embodiment of FIG. 23 FIG. 25 is a block diagram of the modem of FIG.

the base station shown in FIG. 3; FIG. 26 shows the signal interface between the channel controller, the modem and the system timing unit in the FIG. 2 vi²⁰ FIG. 27 shows the signal interface between the modem and the radio frequency unit of FIG. 2 as shown in FIG. 3; FIG. 28 is a block diagram of antenna interface circuit²⁵ shown in FIG. 3, and FIG. 29 is a block diagram of the antenna interface circuit of the FIG. 2.

In addition to the usual abbreviations within the applicable technical field, the description will be used ³⁰ the abbreviations below:

ADPCM Adaptive Differential Pulse Code Modulation =

<td colspan="3">Adaptive DPKM,</td>

<td colspan="2">BCC</td><td rowspan="2">Baseband Control Channel = Baseband Control Channel channel, Binary Phase Shift Keying Modulation = Binary</td>

<td></td><td>BPSK</td>

<td> 5</td><td></td><td>phase modulation,</td>

<td></td><td>CCU</td><td>Channel Control Unit = channel controller,</td>

<td></td><td>CODEC</td><td>Combined Coder and Decoder = Combined the coder / decoder,</td>

<td></td><td>DID</td><td>Direct Inward Dial,</td>

<td> 10</td><td>DMA</td><td>Direct Memory Access = direct storage access,</td>

<td></td><td>DPSK</td><td>Differential Phase Shift Keying Modulation = Differential Phase Shift Modulation,</td>

<td></td><td>DTMF</td><td>Dual Tone Multi-Frequency signaling scheme = 2 tone multifrequency signaling,</td>

<td> 15</td><td>ECL</td><td>Emitter-coupled Logic = emitter logic,</td>

<td></td><td>FCC</td><td>United States Federal Communications Commission,</td>

<td></td><td>FIR</td><td>Finite-Duration Impulse-Response filter = FIR filter,</td>

<td> 20</td><td>IN</td><td>In-phase = i-phase,</td>

<td></td><td>kbps</td><td>kilobits per second,</td>

<td></td><td>MDPSK</td><td>Multi-phase Differential Phase Shift Keying Modulation = multi-phase DPSK,</td>

<td></td><td>MPM</td><td>Message Processing Module = message processing</td>

<td> 25</td><td></td><td>treatment module,</td>

<td></td><td>OCXO</td><td>Oven Controlled Crystal Oscillator = oven-controlled crystal oscillator,</td>

<td></td><td>PBX</td><td>Private Branch Exchange or Automatic Switch = Private subscriber center or automatic switcher</td>

<td> 30</td><td>PCM</td><td>Pulsed Coded Modulation = PKM,</td>

<td></td><td>PSN</td><td>Public Switched Network = public network,</td>

<td></td><td>PSTN</td><td>Public Switched Telephone Network (public telephone network) or other telephone company (eg Telco),</td>

<td> 35</td><td>Q</td><td>Quadrature = interphase,</td>

<td></td><td>QPSK</td><td>Quadrature Phase Shift Keying Modulation =</td>

cross-phase modulation,

RBTG Ringback Tone Generator - callback tone generator,

RCC Radio Control Channel = radio control channel,

RELP Residual Excited Linear Prediction,

RFU Radio Frequency Unit = radio frequency unit,

RPU Remote-Connection Processor Unit = remote connection processor unit,

RX Receive = receive,

SIN Subscriber Identification Number - subscriber identification number,

SLIC Subscriber Loop Interface Circuit = subscriber line interface circuit,

STIMU System Timing Unit = system timing unit,

STU Subscriber Station Telephone Interface Unit * subscriber interface unit,

SUBTU Subscriber Timing Unit = subscriber timing unit

TDM Time Division Multiplexing = time multiplexing,

TDMA Time Division Multiple Access = time multiplexed approach,

Telco Telephone Company,

TX Transmit = send,

UTX-250 Coupler with processing and interface; may be a PBX, your Unique Word = unique word,

VCU Voice Codec Unit = voice codecs / decoder,

VCXO Voltage Controller Crystal Oscillator = voltage controlled crystal oscillator.

It should be noted at the outset that while the specification refers to a particular frequency band e.g. 454-460 MHz, the invention can also be applied at least across the entire VHF, UHF and SHF band.

The FIG. 1, in accordance with the invention, allows telephone connections to be established to local subscribers using UHF radio connection between subscriber stations 10 and a base station 11. The base station 11 establishes connections directly between the radio-equipped subscriber stations 10 and is connected to a telephone company's central 12 for calls to and from points outside the system. The system shown, for example, operates on a couple of carrier frequency channels in the band from 454 MHz to 460 MHz. This frequency band comprises 26 specified channels spaced at 25 kHz with an allowable bandwidth of 20 kHz. The distance between transmit and receive channels is 5 MHz with the center frequency of the lower of the two frequencies assigned to transmissions from the base station. As mentioned earlier, the system can also operate on others

UHF channel pairs.

The transmission from the base station to the subscriber station, ie. over the transmit channel, takes place by time multiplexing. The transmission from the subscriber station to the base station, ie. over the receiving channel, takes place at

TDMA, ie time-multiplexed approach.

All systems are designed to be compatible with 47 CFR FCC rules 21, 22 and 90, as well as with other relevant rules.

The communication between the base station 11 and the subscriber stations 10 takes place digitally by filtered multiphase DPSK modulation on full duplex channels in the frequency band 454-460 MHz with channel spacing of 25 kHz, which meets the bandwidth requirement of 20 kHz, cf. FCC rules 21, 22 and 90 (e.g. 21.105, 22.105 and

90209). The system can also be used in other frequency bands and with other channel distances within the usable parts of the VHF, UHF and SHF bands.

The symbol amount over each 25 kHz FCC channel is 16 ksps in each direction. The voice transmission is done using 16-level PSK modulation and digitization / speech signal is done at a coding rate of 14.6 kbps. Alternatively, there may be binary phase shift modulation (2 levels) and cross-phase shift modulation (4 levels). A mixture of different modulation levels can be used simultaneously on the same channel. For time multiplexing, the system secures a call for each phase multiple of two at the 14.6 kbps rate (4 stages give two calls, 16 stages give 4 calls, etc.) or more for lower rates. Of course, this should only be considered as an example, as the table below provides for the possibility of many different combinations of modem bits / symbols or phases and codes / decoder measures:

Table I road conversations or

Duplex circuits with codecs on:

<td></td><td>Modulation Phases</td><td>14.4 kbps</td><td>6.4 kbps</td><td>2.4 kbps</td>

<td> 20</td><td> 4</td><td> 2</td><td> 4</td><td> 8</td>

<td></td><td> 8</td><td> 3</td><td> 6</td><td> 12</td>

<td></td><td> 16</td><td> 4</td><td> 8</td><td> 16</td>

<td></td><td> 32</td><td> 5</td><td> 10</td><td>²⁰</td>

<td></td><td> 64</td><td> 6</td><td> 12</td><td> 24</td>

<td> 25</td><td> 128</td><td> 7</td><td> 14</td><td> 28</td>

<td></td><td colspan="3">The base station is capable of</td><td>broadcast and</td>

take over any of the available FCC 25 kHz frequency channels in the 454-460 MHz band where the channels can be selected, or optionally over all these channels. The channel frequency selection for each voice channel is done automatically at the base station, one at a time, but a console at the base station allows an operator to disable this feature. The base station can operate with a transmit power of, for example, 100 W for each frequency channel.

/

The base station performs modulation control as well as allocation of time interval and frequency channel to the subscriber stations. In addition, the base station performs adaptive power control over the subscriber stations to minimize sequential time interval differences and the interference between adjacent channels.

The base station switches between the telephone lines' connection lines and between the time-multiplexed intervals over the selected channel, preferably using a digital coupler, which can, however, be replaced by an analog coupler.

The base station creates triple spatial diversity (triple spatial diversity) for the receiving channels.

The subscriber station is capable of operating with three branch diversity. For example, the transmit power is adjusted between 0.1 and 25 W, but can be adjusted to other power values. While the telecommunications through the subscriber station is perceived as real-time full-duplex, the radio frequency system operates at half-duplex at an appropriate time multiplexing.

The subscriber station has an interface with any telephone device for voice communications, but the phone may be built into the system. Furthermore, data connection is possible with conventional 25-pin connector RS-232C for 9600 Baud transmission between subscribers. The base station and the subscriber station can be supplied in any way with internal or external source.

FIG. 2 is a block diagram of one embodiment of the base station simultaneously serving two pairs of radio frequency transmitting and receiving channels. Each channel can carry up to four simultaneous telephone connections. In the preferred embodiment, there may be many transmitting and receiving channel pairs and there are several time intervals in each channel.

One of these available time intervals is required for a radio control channel.

The connections between the telephone network and the subscriber stations are established and maintained in the subscriber center

15 in the base station. This subscriber center 15 may, for example, be of the standard type UTX-250 from the company United Technologies Building Systems Group. Many of the existing features of such a subscriber center are used to control the interfaces with the public network of the system of the invention. The subscriber exchange 15 also converts the voice signal from the telephone network to 64 kbps μ-expanded digital sampler by PKM modulation. From now on, the speech information is processed in digital format over everything in the base station and the subscriber stations and up to the interface circuit connected to the subscriber telephone or to the extent the subscriber's sending and receiving equipment allows.

The voice information in digital form from the subscriber center 15 is then processed in a voice compression system, namely a codec 16 which reduces the speech information rate from 64 kbps to approx. 14, 6 kbps or less. The encoder / decoder 16 uses either a RELP algorithm or an SBC encoder / decoder to perform this clock compression. Typically, there are four encoder / decoders 16 in a single speech encoder / decoder unit 17 to perform speech compression at least four time intervals in each frequency channel. Each encoder / decoder unit 17 can process at least four full duplex voice connections for both the transmit channel and the receive channel in each channel pair. The connections through the switchgear 15 determine which call is processed by which encoder / decoder unit 17, or by which codec 16 in the selected unit 17. The circuits in the individual encoder / decoder units 17 are determined so that a call at a given frequency and assigned a given interval of the base station is always processed by the same codec 16 in the unit.

The individual encoder / decoder units 17 are connected to a channel control unit 18 which controls the time multiplexing function (TDMA) and also acts as a link level protocol processor. The channel controllers 18 receive the output signals of said codec 16 in the respective unit 17 and transmit the data in the correct time interval and format to a modem 19. The channel controller 18 determines the modulation levels to be used for broadcast (e.g., 2, 4 or 16 levels PSK modulation) under the control of a remote connection processor unit 20. The individual channel controller 18 also processes control information for communication to the subscriber stations through the radio control channel time interval and initial control15 bit in the speech channels. Each channel pair comprises a serial combination of a VCU 17, a CCU 18, and a modem 19.

After appropriate formatting, the data from the respective channel controller 18 is transmitted to the corresponding modem 19 at a rate of 16 ksps. Modem 19 takes over these synchronous symbols and performs Graycode formatting through multi-level phase shift modulation. The output of modem 19 is a modulated medium frequency signal. This signal is transmitted to the RF / IF radio frequency unit 21 which converts the medium frequency signal to a

UHF signal in the 450 MHz band. The control signals for modem 19 and radio frequency unit 21 are supplied by the corresponding channel control unit 18 which is itself subject to main control from the RPU unit 20. The UHF signal is amplified in power amplifiers in the radio frequency unit 21 and outputs through the antenna interface unit 22 and a transmit antenna 23.

The receive function of the base station is essentially the opposite of the transmit function. Units 21, 19, 18, 17 and 15 operate full duplexes.

The remote control processor unit 20 is the central control processor which provides for transmitting the connection data and control signals to the channel controller. RPU 20 is a general-purpose computer, e.g. a 6800 microprocessor that manages the system functions and controls for call, interrupt, and maintenance. RPU

20 also communicates with a call processor 24 in the exchange 15 to control the connections between the individual codec 16 and the telephone lines of a switching matrix 25 in the exchange 15.

The subscriber stations are relatively small entities that are with the users of the system. The subscriber station connects the user's conventional telephone and / or data terminal or integrated acoustic transmitter / receiver with the base station through the UHF channel. The function of the subscriber station is substantially the same as that of the base station, but while the base station can simultaneously operate on one or more frequency channels, each capable of carrying multiple voice connections, the subscriber station will normally operate only on a single frequency at a time.

FIG. 3 shows a block diagram of a subscriber station. The functional breakdown is almost the same as at the base station (Fig. 2). The subscriber station includes a subscriber interface device with which the user engages. The corresponding function of the base station is performed by

PBX module. The STU unit in the subscriber station also performs all the subscriber station control functions in the same way as the RPU unit in the base station. Throughout the system, the base station and the subscriber stations operate according to the master / slave principle. The STU can interface with an external instrument or transmit and receive acoustically.

After the data flow through the subscriber station, the user's speech or data information is first processed in the STU unit 27. The voice signal from the subscriber's telephone is received and digitized in the voice code / decoder unit

28th The digitized voice signals have the same format as the format used by the central station 15 in the base station. The subscriber station comprises a VCU 28, a CCU 29, a modem 30a, and an RFU 31a, which perform the same functions as the corresponding units in the above with reference to FIG. 2 base station described. One difference is that the base station operation is usually limited to a single voice channel at a time. The subscriber station operates at half duplex, transmits in one part of the TDMA raster and receives in another part of the TDM10 raster. With a screen size of 45 ms, the subscriber station's half-duplex mode is transparent to the user, thus continuously hearing the speech signal from the counterpart at the other end of the connection. To enable more than one subscriber call can STU 27, VCU 28 and modem

30a is duplicated.

Half-duplex operation at the subscriber station enables more efficient utilization of the subscriber station's hardware. The subscriber station VCU and CCU units operate essentially the same as the base station, at least 20 in terms of processing the voice data. However, modem 30a is arranged to operate in half-duplex mode so that either the receiving portion or the transmitting portion of the modem is used, but not both at the same time. The major saving here lies in the fact that RFU 31a only needs to operate in half duplex mode. This saves power since the radio frequency power amplifier is only active for half the time. The radio frequency transmitting antenna 32a may also be coupled to operate as a second receiving antenna in the receiving portions of the raster. This can be done using an antenna coupler. Furthermore, no duplexes are required.

Each subscriber station also includes a three-modem diversity network and a diversity combination circuit 33 which collects the received, demodulated information from each demodu la tord in the three modems 30a,

Ty · - ·

30b and 30c, combining this information to form a best-guess symbol flow applied to the channel controller 29 for processing. The demodulator circuits of the three modems 30a, 30b and 30c are connected to separate receiving radio frequency units 31a, 31b, 31c and thus to separate antennas 32a, 32b and 32c.

In the base station, there are three receiving antennas 34a, 34b and 34c, which are spaced apart to produce uncorrelated, spatially different signals processed by a diversity network. This diversity network is transparent to the CCU function and can therefore be replaced by single modem function at any time when the diversity function is not required.

The base station also includes a spatial diversity network for each pair of transmit and receive channels. This diversity network is not shown, but the diagram of the base station of FIG. 2 is the same as the diagram of the subscriber station of FIG. 3, which shows the connection of the diversity network for a single transmitting and receiving channel pair. Thus, each transmit and receive channel pair in the base station includes three demodulator parts and a modem connected to the diversity combination circuit as shown in FIG.

3.

Accurate time synchronization between the base station and the subscriber stations is critical in the global system. The base station provides time management of the entire system. All subscriber stations in a given system must be synchronized with the master clock in terms of frequency, symbol timing and raster timing.

The base station includes a system timing unit

35, which produces a very accurate 80 MHz clock signal. This 80 MHz clock signal is divided to produce a 16 kHz clock signal and a raster marking signal of

22.222 Hz (duration 45 ms). The entire base station's sendingDK 175353 B1 timing is based on these three synchronous master references. The 80 MHz clock signal is used by the modems 19 and the radio frequency units 21 to precisely define the intermediate frequency and radio frequency. The 16 kHz clock signal defines the symbol frequency for transmissions at all frequencies of the base station. The 45 ms marking signal is used to mark the first symbol in a new grid. This marking is active for a period of one symbol time (62.5 ys corresponding to 1/16000 Hz). All frequency channels in the base station use the same time reference for transmission. The three timing signals (80 MHz, 16 kHz and marking of start of raster (SOF)) are supplied to individual modems 19 in the base station. The modem 19 distributes the appropriate clock signals to the channel control unit 18 and the radio frequency unit 21 in the single series-connected transmit and receive channel pairs. The channel controller 18 uses the 16 kHz signal and the SOF marker signal to timing the transmission of the speech and control symbols according to the particular raster structure at that frequency.

Ideally, receiving timing in the base station will be identical to the base station's transmission timing. Ie that the SOF marker signal and the symbol rate signal should be exactly synchronized between the transmit and receive signals.

However, in the transmission from a subscriber station one cannot expect perfect synchronization, the reception timing of the modem 19 in the base station must be adapted to the arriving symbols from the subscriber station. This is required in order for the sampling period of the modem 19's take-off function at the base station to create the best rating of the symbol received from the subscriber station. An elastic buffer in the channel controller 18 interfaces with the modem 19 and serves to compensate for such small time delays.

The subscriber stations throughout the system synchronize their rate with the master rate from the base station. This synchronization is achieved in a multi-step procedure whereby the subscriber station captures the base station's time reference by means of the RCC messages from the base station. This procedure is described in more detail below.

After the subscriber station has captured the time reference from the base station, a tracking algorithm ensures in the demodulator portions of the subscriber station modems 30a, 30b and 30c to maintain precise reception timing for the subscriber station. The subscriber station shifts its transit missions back to the base station a short time interval rather than offsetting the transmission delay due to the distance to the subscriber station. This method causes the transmissions from all subscriber stations in the base station to be received with the correct input phase.

The system timing unit 35 creates a time base for all transmissions from the base station. The system timing unit 35 comprises a thermostatically controlled crystal oscillator operating at a fixed frequency of 80 MHz —9 with high precision (3 · 10). This basic clock frequency is divided by 5000 in the system timing unit 35 to produce the 16 kHz symbol clock signal, and again by 720 to create a marker signal indicating the beginning of a raster, ie. SOF signal. These three time references are added to a buffer and delivered to all the base station modems.

The subscriber timing unit, not shown in FIG. 3, at the subscriber station produces an 80 MHz clock signal, a 16 kHz symbol clock signal, and a 45 ms raster marking signal. These signals are identical to the signals in the base station system timing unit, except that the 16 kHz clock signal is used for timing at symbol reception in the subscriber station. The 16 kHz signal is used for timing of broadcast in the base station. Transmission timing in the subscriber35 station is obtained with a delayed version of subscriber DK 175353 B1 station receiving timing. The delay is variable and determined by the distance calculation made between the base station and the subscriber station.

The timing reference signal for the subscriber station is generated by a voltage controlled crystal oscillator operating at a nominal frequency of 80 MHz. The current frequency is adjusted by the subscriber station modem to be frequency locked to the base station timing reference received by the subscriber station radio frequency unit.

Protocols.

The following protocols determine the procedures for system management, collision avoidance and call signaling in the system, as well as for determining the structure of broadcast grids. Referring to the components of the system, reference is made to those above with reference to FIG. 2 components of the base station, unless otherwise stated.

The system uses full-duplex channels with 20 kHz bandwidth in the 450 MHz band, with a frequency range of 25 kHz, and the system carries multiple simultaneous calls on each channel. Each full-duplex channel comprises a receive frequency and a transmit frequency with 5 'MHz separation. The lower frequency in each channel is assigned the base station for the trans25 mission and is called the forward frequency. The upper frequency in each channel, so-called return frequency, is assigned to the subscriber stations for transmission. The base station thus broadcasts on the forward frequency and receives on the return frequency. The opposite applies to the subscriber stations.

The ability of the system to create a spectrally effective method of transmitting over multiple voice channels at a single frequency depends primarily on the modem operation. Modem 19 must operate so that it creates

3.2 bit / Hz efficiency when operating with differential 35 phase shift modulation at 16 levels with symbol frequency of 16 ksps.

The modem 19 is a coupling adapted to convert symbols of 1, 2, 4 or more bits from the channel controller 18 into a phase modulated medium frequency carrier used for transmission and to perform the reverse conversion on the receiving side. The entire control process for raster timing and mode selection is performed by channel controller 18. An interface between channel controller 18 and modem 19 may consist of two unidirectional, synchronous (16 ksps) 4 bit data buses (Tx and

Rx). In addition, an 8 bit status / control bus serves to provide the modem with control information and to report status information from the modem to the channel controller

18th The modem 19 also supplies the channel controller 18 with the 16 kHz master clock signal. In the base station, this clock signal is received from the master oscillator in the system timing unit 35, with which the entire base station and thus the whole system is synchronized. In the subscriber station, this clock signal is derived from the symbols arriving from the base station. Therefore, all transmissions have reference to the time base in the base station. A major feature of the subscriber modem operation is to synchronize the local clock signal with the base station's time reference by decoding the timing of the received symbols.

The modem part of the transmitter modem uses a digital one

FIR filter to create a digital representation of the waveform used to modulate the radio frequency carrier. The resulting digital stream is converted to analog form and mixed with a transmit intermediate frequency of 20.2 MHz. The signal is then applied to the radio frequency unit for filtering, further conversion to radio frequency and amplification prior to transmission.

The receiver's demodulator portion acquires the received medium frequency signal from the radio frequency unit

21 at the medium frequency of 20 MHz. This signal downconDK 175353 B1 is converted to baseband and then digitized using an analog / digital converter. The resulting digital samples are processed in a microprocessor based signal processing unit. This function algo5 rhythms for filter equalization and synchronization on the arriving samples and then demodulate the phase shift modulated signal to generate a symbol flow of 16 ksps. The signal processing unit also operates in self-training mode which is used to notify the processing unit of inaccuracies in the analog filters used for the received current. After training the signal processing unit, the demodulator's digital equalization process compensates for the effect of such inaccuracies in the analog filter components on the upcoming samples. This technique favors the use of cheaper, lower-tolerance analog components and helps enable the system to demodulate weak or noisy signals.

In receiving mode, the modems demodulated by the modem are passed to the channel controller 18 at the given symbol rate. Modem 19 creates the timing associated with this symbol stream. Both the base station and the subscriber stations derive receive function timing from the arriving received signals.

A more detailed description and specification of the modem features and characteristics can be found later in the description with reference to FIG. 25th

This basic TDM / TDMA channel structure for each subscriber creates a total capacity of 16 kbps for each call in both directions. This channel capacity requires 1.43 kbps in each direction for the initial control and demodulation preamble. The VCU therefore operates at a fixed data rate of 14.57 kbps. This is equivalent to 328 bits per second. code / decode raster period, define35 right as one half of the modem raster period, ie. 22.5 ms.

In order to carry more calls per channel is each channel by time multiplexing divided into intervals (slots). These ranges define the grid format of the system. The length of the system grid covers a predetermined constant number of symbols. The system grid duration has been optimized taking into account the speech coding rate and the number of capture symbols required by modem 19 at the beginning of each burst. The number of intervals within the system grid depends on the modulation level of the channel. For example, if the channel is QPSK modulation, the system grid has two intervals per channel. raster. By increasing the modulation level of the channel, the number of coded information bits per symbol, which is why the channel's data volume is increased. For 16-level DPSK modulation, the system grid is divided into four intervals, each carrying the voice data for one call. It is important to note here that even at higher modulation levels, the required number of symbol time intervals for modem synchronization remains constant.

The format of the system grid ensures that the modem 19 of the subscriber stations never has to operate full duplex, ie. to send and receive simultaneously. Therefore, the intervals of the forward and return frequencies are temporally offset by at least one interval.

The system grid for the system has a duration of 45 ms. The symbol transmission frequency is 16 ksps. Each symbol is emitted at the same time duration, namely 1/16000 of a second, i.e. 62.5 ps. There are 720 symbols per day. raster, which symbols from the beginning of the raster are numbered from 0-719. These 720 symbols may consist of 1, 2, or 4 bits of information each, corresponding to modulation of 2, 4, or 16 phase levels.

The 45 ms system raster time is again divided into two or four sub-intervals depending on the modulation forDK 175353 B1 fed for the intervals in the raster. There can be three different types of intervals: 1) radio control channel (RCC), 2) 4-level voice channel and 3) 16-level voice channel. The RCC channel always broadcasts in binary modulation, ie. two-phase modulation. Each interval for the RCC channel and the 16-level voice channel requires 180 symbols to be output, ie. a quarter of a system grid period. Since the 16 voice channel transmits 4 information bits 4 per symbol, ie 2 = 16 phases, transmits the 16 voice channel

720 information bit per raster. This corresponds to a bit rate of 16 kbps. Some of these bits are used as an introduction to the modem or for control, leading to a speech bit rate of 14.57 kbps. The interval for the 4-voice channel requires 360 symbols to be output, corresponding to half the system grid period. Each symbol in an interval of this kind comprises 1-4 differential phases, so that 2 bits are output per second. symbol (2 = 4 phases). The resulting bit rate is 16 kbps, ie. the same as for the 16 voice channel. The same number of bits (not symbols) is reserved as the introduction for the modem and for control, so that the voice information frequency is 14.57 kbps, ie. the same as in the 16-voice channel range.

The system grid for any given frequency 25 channel may consist of any combination of these three types of intervals, under the following five conditions:

First A maximum number of symbols (720) is issued in each system grid. Combinations of the three types of intervals can be used at a given frequency to achieve this. In the case where full channel capacity is not recorded in raster transmission from the base station, ie. if less than 720 symbols are transmitted in a grid, zero symbols are entered to fill the grid capacity up to 720 symbols. A zero symbol is defined as a symbol in which no power is emitted.

2nd In a multi-frequency base station, there is only a single frequency that includes an RCC interval. At any one time, only a single RCC can operate throughout the system. The frequency at which this RCC operates is established by a system initialization parameter and is only changed when this frequency channel is rendered inaccessible for some reason. The RCC interval is always assigned the first 180 symbols in the system grid and is designated as O interval.

Third The frequency of a base station can operate in constant transmission mode. The subscriber station broadcasts for no more than half of the total raster time. The subscriber station when transmitting a call broadcasts only less than 25% of the raster when operating in RCC or 16-voice dialing mode. The subscriber station will broadcast less than 50% of the grid when operating in 4-voice channel mode. A subscriber station can only broadcast at one interval in any raster when transmitting a single call.

4th All 4-voice channels must begin transmission on symbol # 0 or 360. This means that either the first half or the second half of a grid may contain a 4-voice channel.

5th The transmissions on the forward frequency and the return frequency are assigned such that for a given interval the return message begins transmission 180 symbols after the transmission of the message on the forward frequency. This excludes for the subscriber station the requirement to broadcast on the return frequency while receiving on the forward frequency.

Under these conditions, up to four calls can be processed on a single frequency if all four calls consist of 16-voice channel format and operate with 14.4 kbps codecs.

The intervals in the system grid are numbered according to their position in the grid structure. The numbering system need not be continuous. When one or more of the intervals in the grid is a 4-digit channel interval, the numbering system skips the second interval period included in the longer 4 interval.

The interval numbering system for transmissions on the return frequency, ie. from the subscriber, is staggered from the numbering for transmission from the base station on the forward frequency. Therefore, a subscriber receiving information in interval 2 of the forward frequency will transmit in interval 2 of the return frequency, which is time-shifted by half of a raster. The following tables 1-5 show possible grid formats and the numbering associated with each interval.

Table 1

Radio control channel with BPSK

Above way Channel;

&lt;-------- System grid = 45 ms --------&gt;

&lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;-11.25-&gt; ms

180

BPSK

180

16-PSK interval # symbol number modulation type

AM HOLE FILTER BIT SYNC STARTUP PATTERN 8 46

RCP function

112 symbol number./

Return Channel:

<td rowspan="2"></td><td colspan="2">&lt;-11.25-&gt; &lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;-ll, 25-&gt; ms</td>

<td> 2 3 0</td><td>1 interval no.</td>

<td> 5</td><td> 180</td><td>180 symbol numbers</td>

<td></td><td>BPSK</td><td>16-PSK modulation type</td>

<td> 10</td><td>RANGE</td><td>1 FILTER</td><td>BIT SYNC</td><td>UW</td><td>RCP</td><td>RANGE 2</td><td>function</td>

<td></td><td></td><td>STARTUP</td><td>PATTERN</td><td></td><td></td><td></td><td></td>

<td></td><td>XX</td><td> 8</td><td> 49</td><td> 8</td><td> 112</td><td>3-XX</td><td>symbol number</td>

= 0/1/2/3

Table 2

Voice channel grid structure 4 levels

&lt;---------- System grid = 45 ms ----------&gt;

&lt;----- 22.5 ms -----&gt; &lt;----- 22.5 ms -----&gt;

360

360 interval number symbol number

FILTER BIT SYNC PASSWORD VCF 0 VCF 1 STARTUP PATTERN 8 18

164

164 function symbol number

Return Channel:

&lt;----- 22.5 ms -----&gt; &lt;

ISLAND

360 —22, 5 ms -----&gt;

360 interval number symbol number

<td></td><td>FILTER</td><td>BIT SYNC</td><td>PASSWORD VCF 0</td><td>VCF 1</td>

<td> 10</td><td>STARTUP</td><td>AGC</td><td></td><td></td>

<td></td><td> 8</td><td> 18</td><td> 6 164</td><td> 164</td>

function symbol number

Table 3

Voice channel grid structure 16 levels

&lt;---------- System grid = 45 ms ----------&gt;

&lt;-11,25-&gt; 0

180 &lt;-11,25-&gt; 1

180 &lt;-11,25-&gt; 2

180 &lt;LL, 25-&gt;

180 ms interval No. symbol number

FILTER BIT SYNC PASSWORD VCF 0 VCF 1 function

STARTUP PATTERNS 25 8 5 symbol number '•' ΤΙΓΊΜΠ gaaBaepg «^

- = 'a; -τ'-_:

Return Channel:

&lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;-ll, 25-&gt;

ms

<td></td><td> 2</td><td> 3</td><td> 0</td><td> 1</td><td>interval no.</td>

<td> 5</td><td> 180</td><td> 180</td><td> 180</td><td> 180</td><td>symbol number</td>

<td>FILTER</td><td colspan="2">BIT SYNC CODE WORD VCF 0 VCF 1</td><td>function</td>

<td>10 STARTUP</td><td>AGC</td><td></td><td></td>

<td> 8</td><td> 5</td><td> 3 82 82</td><td>symbol number</td>

<td></td><td></td><td>Table 4</td><td></td>

<td colspan="2">Grid structure at</td><td>mixed modulation:</td><td>2/16-PSK and 4-PSK</td>

Above way Channel:

&lt;------- System grid = 45 ms -------&gt;

<td rowspan="2"> 20</td><td colspan="3">&lt;-ll, 25-&gt; &lt;-ll, 25-&gt; &lt;----- 22.5 -----&gt;</td><td rowspan="2">ms interval no. module station type symbol number</td>

<td colspan="2">0 1 2/16-PSK 16-PSK 180 180</td><td>2 4-PSK 360</td>

<td></td><td>Return Channel: &lt;----- 22.5 -----&gt; &lt;</td><td> -11,25-</td><td>&gt; &lt;-ll, 25-&gt;</td><td>ms</td>

<td></td><td> 2</td><td> 0</td><td> 1</td><td>interval no.</td>

<td> 25</td><td>4-PSK</td><td colspan="2">2/16-PSK 16-PSK</td><td>modulation type</td>

<td></td><td> 360</td><td> 180</td><td> 180</td><td>symbol number</td>

For the definition of each interval symbol, cf. 2-1 to 6-3.

• 32

Table 5

<td colspan="3">Mixed modulation: 4-PSK and 16-PSK</td>

<td>Above way Channel:</td><td></td><td></td>

<td> &lt;-----22,5-----&gt; &lt;-</td><td>ll, 25-&gt; &lt;-ll, 25-&gt;</td><td>ms</td>

<td> 0</td><td> 2 3</td><td>interval no.</td>

<td>4-PSK</td><td>16-PSK 16-PSK</td><td>modulation type</td>

<td> 360</td><td> 180 180</td><td>symbol number</td>

<td>Return Channel:</td><td></td><td></td>

<td>&lt;-ll, 25-&gt; &lt;-ll, 25-&gt;</td><td> &lt;-----22, 5-----&gt;</td><td>ms</td>

<td> 2 3</td><td> 0</td><td>interval no.</td>

<td>16-PSK 16-PSK</td><td>4-PSK</td><td>modulation type</td>

<td> 180 180</td><td> 360</td><td>symbol number</td>

<td colspan="2">Table 3 shows the interval structure</td><td>for a 16-voice</td>

channel with 180 symbols. The first 8 symbols in this type of interval are filter startup bits. The filter start period, which exists at the beginning of each type of interval, is the time at which no energy is emitted to give the receiving portion of modem 19 of the receiver time to clean its filters before the new interval arrives.

After the filter startup, the bit synchronization period comes, during which a degenerated 16 level is emitted simulating an alternating BPSK signal. The receiving portion of modem 19 uses this field to establish phase reference for the transmitter portion of modem 19.

Then a 12-bit password is used to synchronize the subscriber with the base station and to exchange control and status information. The passwords are used to indicate the current state of the connection, connection quality and power and timing adjustment. Each password is encoded to 10 bits using a single 35 error correction and double error detection Hamming code. The channel controller UK 175353 B1 unit 18 determines the gain and loss of synchronization by tracking the number of successive passwords received correctly or incorrectly, and transmits synchronization change message to the remote connection processor 20 at the base station. At the subscriber station, the channel controller 29 transmits the synchronization changes to the subscriber interface unit 27.

The Hamming code adds five parity bits to five information bits to create a 10-bit password. Each parity bit is calculated by modulo-2 addition of all bits in positions within the codeword containing the bit represented by the parity bit. As the password is transmitted with all bits lying next to each other followed by all the parity bits, by arranging the parity bits in positions within the word with just a single bit in the position represented by the bit, and by placing the data bits in the other positions, the password can be visualized. in the following way:

bit position:

info: Pl P2 Dl P3 D2 D3 D4 P4 D5 P5

P = parity bit

<td> 25</td><td>D = data bit</td>

<td></td><td>PI = D1 + D2 + D4 + D5</td>

<td></td><td>P2 = D1 + D3 + D4</td>

<td> 30</td><td>P3 = D2 + D3 + D4 P4 = D5 P5 = total</td>

When a password is received, the parity bits are calculated from the received bit of data and compared to the received parity bit. If the calculated total parity 175353 B1 bit bit differs from the received total bit, exclusive or processing of the calculated parity bit and the received bit is performed to obtain the address of the bit where there is an error. If the calculated total bit and the received total bit match, but not the other four bits, two errors are detected. If all parity bits match, the data is received correctly.

The rest of the interval contains two speech codecs on each 328 bit of information.

Table 2 shows the symbol structure of the 4-voice channel.

The structure is almost the same as for the 16-voice channel. The differences are because the allocation of certain symbols depends on a given number of symbols required for each interval as initial symbols, while other bits15 allocations relate to a fixed number of bits.

The radio control channel has the dual function of allowing the subscriber stations to receive system timing from the base station and to create out-of-band signaling between the base station and the subscriber stations.

For the radio control channel, there is the same range format for the forward channel and the return channel, except for subsequent fields. The first 8 symbols in a control interval emitted from the base station (forward channel) contain an amplitude modulation hole (AM Hole), ie. a period in which no energy is emitted. This gap is used to clearly identify at the subscriber stations the control channel. At the beginning and end of the return channel control interval, there are a few additional symbols to take into account the fact that the subscriber stations could be off-set with a few symbols in terms of timing.

All intervals contain 8 symbols for the '' 0 transmission and the filter startup field, which gives mode35 with the ability to clean the receiving filters as prefixDK 175353 for the new interval. The next field in the range is a fixed bit synchronization pattern. The transmitted pattern is an alternating BPSK signal. The receiving modem uses this field to establish a phase reference and perform frequency locking to the transmit modem.

Channel controller 18 is constantly looking for a unique word (UW), which is a sequence of 8 symbols, for identifying an oncoming RCC message. The base station channel controller 18 must perform an exhaustive search for a valid RCC message at each RCC interval. It performs this task by scanning for the unique word in a window of ± 3 symbols around the nominal location of this word, based on master system timing. The search algorithm starts with the nominal word placement and switches one symbol to the right and left until it 1) finds the UW pattern, and

2) verify that there is a proper RCC checksum. The search is completed as soon as conditions 1) and 2) are fulfilled or when all options have been searched under 20. The shift information, RCC message and power information are sent to the RPU 20 after a successful search.

The subscriber station channel controller 29, when receiving RCC data, may be in one or the other of two modes: raster search or monitor. Grid search mode is used to capture receiver grid timing from the arriving RCC data and is automatically triggered when RCC synchronization is lost. Monitor mode is used when the receiver master synchronization is established.

When in raster search mode, the CCU 29 of the subscriber station must perform an exhaustive search for a valid RCC message immediately after receiving an RCC interval at the subscriber station. Similar to the base station CCU unit

18 CCU 29 performs this task by scanning for the unique word in a window of ± 3 symbols around the nominal UW location, based on timing derived from the modem's detection of AM hole. The search algorithm begins with the nominal UW position and switches one symbol to the right and left until 1) it finds the UW pattern and 2) verifies that the RCC checksum is correct. The search is terminated as soon as conditions 1) and

2) are fulfilled or when all possibilities have been explored. The shift information for a successful search is used to adjust the receiver control markings produced by the channel controller. The capture ends when conditions 1) and 2) are met for three successive rasters with the unique word UW located in its nominal position. The subscriber interface unit 27 is notified of raster capture when this occurs. In grid search mode, the RCC messages are not passed to the subscriber interface unit 27.

When the raster capture is done, the subscriber station channel controller 29 switches to monitor20 mode. Only the nominal position of the UWord is checked to exclude the possibility of incorrect UW capture. If the UW word is not defective in five successive rasters, the channel is declared out of sync and switched to raster search mode (this shift is normally very unlikely, as the system performance is unacceptable in that case). The subscriber interface unit 27 is notified of this out-of-sync state. In monitor mode, the RCC messages having the correct checksum and correct subscriber identification number (SIN) are passed to the subscriber interface unit 27.

The rest of the interval is used to exchange information between the base station and the subscriber stations. The data section consists of twelve bytes. The first 8 data bits include a link field that carries information regarding the system's status, collision, detection and reservation.

The purpose of the link level protocol is to detect incorrect messages over the radio control channel. The Link protocol also resolves disputes during the RCC interval.

The link field also includes idle transmission bit (idle transmission), system busy (system busy), collision, transmission detected and slot

<td>reservation.</td><td>These</td>

<td>nalstyreenhed</td><td> 18</td>

<td>nalstyreenhed</td><td> 29.</td>

<td>Bitten</td><td>for</td>

<td>because station</td><td>for</td>

and is read by the subscriber station's kaidle transmission set up by message. When a subscriber unit receives an interval of this bit, it performs the usual synchronization and error checking operations, but it does not transmit the message to the respective RPUs 20 or STUs 27 when the message is received error-free.

The system busy bit indicates that all voice channels have been allocated and that new calls should not be attempted (for a certain amount of time).

The collision bit resolves conflict between two or more subscriber stations that try to broadcast within the same control interval.

The bit detected for transmission indicates that the base station has detected a transmission over the return channel.

The interval reservation bit causes the next interval over the return channel to be reserved.

The remainder of the data section is used for addressing and exchanging information during the establishment and degradation procedure. After the data section, a cyclic 16 bit redundancy check (CRC) comes over the unique word and data sections in the interval. CRC is used to detect errors that occur during the transmission of the RCC messages. The CRC algorithm includes dividing a block of data with a predetermined bit sequence and transmitting the remainder of this split as part of the data block. The polynomial for generating CRC checks has the following form:

P (x) = 1 + x⁵ + x¹² + x¹⁶ (1)

When the CRC checks for a received message, the message is not passed from the CCU unit 18 to the RPU unit 21 at the base station, or from the CCU unit 29 to the STU unit 27 at the subscriber station.

The moment a subscriber station connects and enters the line, it must capture system timing and synchronization from the base station. This capture fo15 is governed by communication through the radio control channel and by a so-called refinement over the voice channel. The system capture takes place at the following events:

First When the subscriber station is turned on (power on), the system is initialized and the subscriber station CCU unit 29 delivers to the demodulator portions of the subscriber station modems 30a, 30b and 30c a series of orders leading to

RCC-capture.

2nd The demodulator portion of the individual modems 30a, 30b and 30c is first placed in training mode. During this time, the modem trains its digital filters in relation to the characteristics of the receiver's analog filters. These analog filters can change over time and depending on the temperature. In training mode, each modem performs an automatic adjustment of the digital filter coefficients to compensate for such variations. After the CCU 29 of the demodulator parts of Modems 30a, 30b and 30c has been notified that the training sequence is complete, the CCU sets the receive frequency at the incorrect RCC frequency. The CCU then orders the modem to capture the RCC frequency and search for the aforementioned / · characteristic interrupt in the amplitude modulation, ie. the previously mentioned AM hole, which is a period of time of 16 symbols, where during the beginning of an RCC transmission no energy is emitted from the base station. All other broadcast interval types involve only an 0 transmission of 8 symbols. The additional 8 symbols of O information at the beginning of an interval burst cause this burst to be uniquely identified as RCC.

Third First, the demodulator parts of modems 30a,

30b and 30c do is to do a rough frequency capture. The received signal is processed in a digital phase lock loop and the voltage controlled crystal oscillator in the subscriber station is set at the base station transmitting frequency. After frequency capture, the modem searches for the AM hole. The modem searches for a sequence of symbols with low amplitude, possibly amplitude equal to 0. When this sequence is detected in a number of rasters, the modem produces an AM strobe signal to initialize the channel controller's raster timing circuit. If no AM-hole sequence is detected, the modem sends back to the channel controller that the RCC capture was unsuccessful. Then the channel controller begins to search for others in the same way. RCC frequencies.

4th After detecting the AM hole, the demodulator parts of modems 30a, 30b and 30c perform a finer capture and begin bit synchronization. The first 60 symbols in the RCC control interval are a fixed bit synchronization pattern that the modem uses to lock to the base station (bit timing) phase. At this point, the RX rate signal in the subscriber station can be used as a symbol rate signal.

5th The subscriber station CCU unit 29 has received a rough symbol timing adjustment via the AM strobe signal from the mo35 demet. After frequency capture and bit synchronization, the CCU examines the data received by the modem and searches for the unique RCC word. This unique word provides the absolute symbol counter reference for the grid. The CCU then adjusts its symbol counters for this reference. The subscriber station is now set and locked to the base station's system timing in both frequency and symbol.

6th The remainder of the system timing capture determines the distance-dependent delay between the base station and the subscriber stations. This delay can be 0-1,2 symbol times (one way) in the system. During call setup, the subscriber station sends a message to the base station over the RCC channel.

7th The base station modem 19 is always looking for new calling subscribers. The corresponding bursts may be delayed 0-3 symbol times relative to the base station master reference for starting raster. At each interval, the demodulator parts of the base station modems 30a, 30b and 30c search for a transmission in an RCC return interval. The entire timing and phase information must be derived during the first part of the interval (preamble), otherwise the interval and associated information is lost. No other option is available while receiving arriving check intervals. The arriving checks are received in the RCC unit according to the Aloha queue principle.

8th In each interval, the base station modem 19 performs a rapid AGC adjustment and bit timing assessment during the first 60 symbols of the interval. The reception signals rate signals are adjusted to compensate for the subscriber station delay depending on the distance. The received data is then delivered to the base station CCU unit 18, which defects the location of the unique word in the stream and determines the total remote delay between the base station and the subscriber station 175353 B1. Modem 19 provides AGC information to CCU 18 to determine the subscriber station transmit power adjustment. The modem 19 also provides the CCU 18 with connection quality and fractional time information. The connection quality is used to determine whether a collision has occurred. Poor connection quality indicates poor signal quality, most likely due to simultaneous broadcast from multiple subscribers during the RCC interval. The estimated fractional time is the value calculated by the modem 19 of the distance dependent delay between the base station and the subscriber stations.

9th The power and delay information is processed by the CCU unit 18 and fed to the RPU unit 20 which transmits this information RCC format and transmits this information to the subscriber station via the RCC control interval. The subscriber station CCU unit 18 decodes this information and makes the necessary adjustments with regard to the transmit power and delay counters both in modem 19 and CCU unit 18. The CCU 18 updates its own broadcast symbol grid counter and updates the modem's clock delay counters.

10th During the call setup for a subscriber station, the base station's RPU 20 makes frequency and interval allocation for the call. This information is transmitted over the radio control channel and the subscriber station CCU 29 adjusts the receive frequency and commands the modem to begin detecting an interval in the voice signal. The AGC, timing, and frequency information is transmitted from

RCC operation for voice channel operation. This is possible since all frequencies in the system are synchronized on the same raster timing reference in the base station.

11th In order to accurately establish the timing of the subscriber station, a fine tuning procedure is implemented in each telephone connection. During this juDK 175353 B1 stering phase, the communication over the voice channel is almost the same as the control channel used BPSK modulation and the messages are in RCC format, but no AM hole is generated from the base station; these new RCC5 messages are only exchanged between CCU monitors 18 and

29th Modem 19 is set in refinement mode in the base station and in outbound control mode in the subscriber station. During the said adjustment phase, the subscriber station CCU 29 produces a message, the largest of which consists of a fixed bit pattern accompanied by a variable portion indicating whether the previously received message has been accepted or rejected. The base station modem 19 notifies timing and power adjustments to the CCU 18 for each received interval. The power adjustments are continuously sent to the subscriber station. The timing adjustments and control information indicating whether refinement mode continues or is completed are issued after a calculation period. The base station CCU unit 18 collects the timing adjustments from modem 30 over 30 rasters, calculates a mean and then sends the adjustment to the subscriber station CCU unit 29. The base station CCU unit 18 then performs another adjustment over 30 rasters, and the result is sent again to the subscriber station CCU unit 29. The alignment phase is completed in the base station CCU unit 18 and the voice connection begins when the deviations in the adjustments received from modem 19 within an acceptable value range of e.g. 1% or when the adjustment periods have taken a maximum time.

During call setup and degradation, subscriber stations communicate with the base station by sending messages over the RCC return interval. The traffic from the subscriber stations trying to access the RCC35 channel can be characterized as stochastic. When a subscriber station wants to send a message to the base station, there must be some kind of control that determines which subscriber station is allowed to send, as it may happen that several subscriber stations are trying to send in the same interval. The aforementioned Aloha buying principle is well suited to a large number of subscribers requiring relatively rare random access to the RCC channel.

This interval-split Aloha queue principle allows subscriber stations to send messages in the designated RCC range completely independently of whether other stations are also trying to send in the same interval. The natural consequence of this independence is that messages from different subscriber stations can be broadcast at the same time and therefore collide with each other. To resolve this collision problem, the system requires the base station to issue a positive acknowledgment (ACK) after correctly receiving the subscriber station message. If this acknowledgment is not received within the maximum time required for transmission and processing delays in both directions (approximately 1-2 rasterized times), the subscriber station may again send the message. Retransmissions may be due to an error in receiving the recognition at the subscriber station.

Generally, the subscriber stations cannot determine the cause of the problem. Therefore, at the subscriber stations, a random delay is selected prior to retransmission of the message in order to avoid repeated collisions with other transmitters that may have been involved in the previous collision.

One complication that arises in the Aloha queue principle is that the channel may become unstable if the retransmission delays are not long enough. When this happens, the channel becomes overloaded with retransmissions and the system throughput drops to zero. A backoff technique minimizes the problem by increasing each subscriber station's random average retransmission delay with successive retransmissions.

The implications in retransmissions due to collision and in controlling access delay stability are that the delays usually have geometric distribution. Therefore, to avoid large variations in delay, it is necessary to allow the duct to operate with a coating substantially less than 36%.

With a coating of 20% or less, more than one retransmission is unlikely due to collision. If you use a random delay of e.g. 8 frames of 45 ms, the total average delay for one retransmission will be 450 ms, with this average delay including 1 frame delay for the original transmission, plus 1 frame delay for recognition, plus the 8 frames mentioned.

To be sure that the occupancy is not more than 20%, we must take into account the average time T between call requests per call. subscriber, the total number of subscribers N and the raster time F for values ​​less than 36%, thereby defining the coating as NF / T. For

F = 45 ms, N = 1000 subscribers and T = 30 minutes, the utilization rate is 1.5%.

At the maximum occupancy of 20% and with a population of 1000 subscribers, each calling on average every half a minute, traffic can be managed with a speed of 45 ms, with access delays of approx. 45 ms when one retransmission is required and with an average access time of approx. 70-80 ms. The price for the much lower average delay is an increased delay variance which, for a coating of 20% or more, should be longer than two retransmissions, namely 1 s.

The Aloha queue principle proves to be suitable for a system with a large population of subscribers requiring relatively rare random access to the control channel and should, for expected population parameters, allow for delays of less than Is. In contrast, polling systems (subscribers are requested to call) and fixed TDMA lead to unacceptable delays.

All phases in the processing of a call, including establishing calls, breaking down calls, and assigning intervals require the exchange of information over the control channel and / or over the control section of the voice interval. A description is then given of the different phases in the processing of calls, both in terms of the processing at the subscriber station and the processing at the base station.

The subscriber identification number SIN of the subscriber station and the rotated digits are two information which in each call from a subscriber station must be included in a CALL REQUEST message to the base station. In case of calls from subscriber station to subscriber station, the user rotates the number in a register in the subscriber station store. The user initiates communication with the base station by pressing the send key or by waiting for a while. Only when the number is complete and stored in the subscriber station is the radio channel used. Therefore, the user can rotate the number slower without seizing any of the radio control channel's bandwidth or time.

FIG. 4 shows the sequence of messages provided at the subscriber stations and the base station to establish a connection between two subscriber stations. The control level link level protocol is used to check the various error states that occur due to channel errors. Furthermore, the messages received by the base station on the return control frequency are automatically recognized in the next control interval on the forward control frequency. The following sections provide a brief description of a message exchange for call setup between two subscriber stations.

When the base station of a subscriber station A receives a CALL REQUEST message over the control channel, it first examines the received SIN signal for errors. If the SIN signal is faulty, the message is discarded.

Without a valid SIN signal, the base station is unable to know who sent the message. If the rotated or entered digits are incorrect or incomplete, the base station over the forward control channel sends a CLEAR INDICATION message to the calling subscriber station A with status information specifying the problem.

If the first attempt is in order and accepted, ie. if the counterparty is not busy, the voice channel is allocated to the calling subscriber A and the base station sends a PAGE signal in the form of an oncoming call message over the forward frequency to the subscriber B. If, after two attempts, subscriber B does not respond to the PAGE signal with a CALL ACCEPTED message or switches to busy through a CLEAR REQUEST message, then the base station for subscriber A sends a CLEAR INDICATION message with status information that the counterparty is busy or not responding to the search ( PAGE).

If the called subscriber B accepts the incoming call, a CALL ACCEPTED message is sent to30 back to the base station and the voice channel is allocated. When the voice channel synchronization is performed, the subscriber station B produces an audible tone that the subscriber B can hear, as well as a RINGBACK tone to the subscriber station A over the voice channel.

When subscriber B picks up, the voice interval control section switches from a sync ring indication to a sync offhook indication and CALL PROGRESS messages are exchanged between the subscriber stations over the voice channel via the base station. At this point, subscriber station B terminates the audible ring tone and disconnects the ring back tone from the voice channel. The connection is now established and voice signals and / or data signals can be exchanged.

An external telephone call is made in the same way as when calling another subscriber station. The subscriber station simply rotates or enters the desired digits, presses the send key and waits. This generates a radio request message to the base station. The base station decides whether to search for another subscriber station or record an external connection line. In this case, an external connection line is recorded and the rotated or entered digits are sent over the connection line. While broadcasting these digits, voice allocation is made for the calling station. When a subscriber station receives the CALL-CONNECT message, it switches frequency and synchronizes itself to the allocated voice channel. When the voice channel is ready, the subscriber station microphone is disconnected from silent mode and connected to the external connection line. Then, the telephone company's central generates all call tones.

All outside calls take a connecting line to the base station. The calling center sends 2-5 digits that uniquely identify the subscriber station's SIN number to the base station over the direct inward dialing (DID) line. If the called subscriber station is not occupied, the base station sends a PAGE MESSAGE over the radio control channel to that subscriber station. There are three possible situations here. 1) The subscriber station accepts the incoming call and the processing takes place as described below. 2) There is no answer. In that case, the base station repeats the search process twice. If, after a given number of increments, the base station still does not receive a response from the subscriber station, a ringback tone is sent to the calling station. 3) The called station is calling (the tube removed) and sends a CLEAR-REQUEST message back over the control channel. In that case, the calling station gets busy tone.

In the case where search is successful, the voice channel generated ringtone of the called subscriber is allocated and an audible ringback tone from the subscriber station is returned to the calling subscriber. When the called subscriber answers the call, ie. when the base station detects a change from the tube to the tube 15, both the ringing tone and the ringback tone are interrupted. From this point on, the voice channel is ready for conversation.

A normal end of a call is done by the subscriber putting the handset on. The base station detects this through the voice channel control section. When it detects this, the base station cancels the channel allocation. This channel must not be used again until the base station finds that the subscriber station loses synchronization on this channel. If the call being interrupted is destined for another subscriber station, a tube of indication is sent to the other subscriber over the voice channel control section. The subscriber stations resynchronize themselves to the RCC transmissions and send CLEAR REQUEST messages to the base station.

A call can also be terminated 5 seconds after the base station has lost radio contact with a subscriber station.

A voice connection may be lost due to fading or channel interference of the destination receiver. If problems arise in the connection, the following conditions are investigated at the subscriber stations and base station 175353 B1: the connection quality value returned from the subscriber or base station receiver is repeatedly below a predetermined threshold value; there are over several successive transmissions detected loss of word synchronization.

The messages from the base station are transmitted over the radio control channel to all the active subscriber stations. The purpose of the broadcast message is to notify all active subscriber stations of changes in the operation of the system, e.g. changing the frequency of the radio control channel or ordering the modems to switch to self-test mode, etc. These messages are not recognized by the subscriber stations.

Remote Controller Processor (RPU)

The RPU acts as the control computer in the base station. It interfaces with the CCUs 18 communicating with the radio equipment and with the PBX 15 as shown in FIG. 2nd

The RPU 20 coordinates the necessary measures for processing radio calls. The RPU 20 exchanges messages with the subscriber stations, with the PBX 15 and with the CCUs 18 for connection establishment and degradation. The call processing functions include allocation and disallocation of the radio channels. The RPU 20 also maintains a database that continuously indicates the status of the system. The database contains information about the status of the equipment, the subscriber stations, connections and radio channels within the system.

A call setup begins when the RPU receives a message either from the PBX call processor 24 in the case of a call received over an external line, or from a subscriber in the case of a call to an external telephone or another abon / · nent. The communication from a subscriber arrives over the radio control channel via a CCU unit 18 in the base station. The RPU 20 allocates a voice channel and exchanges messages with the subscriber station, the PBX 15, and

CCU unit 18 for establishing the connection.

A connection breakdown begins with receiving a message indicating that the tube has been purged from the PBX device 15 or a subscriber, or from the CCU device 18 indicating that the synchronization over the radio channel has been lost. The RPU unit notifies CCU unit 18 and PBX unit 15 to disconnect and the RCC channel is disassociated.

The RPU device software performs the following functions:

First Processing of subscriber, CCU and PBX messages regarding call setup, call breakdown, and channel allocation.

2nd Initializing and maintaining a write / 20 read database for the system.

Third System console management to allow system queries and manual system management.

4th Controlling BCC Interface by Controlling BCC Communication's Protocol by Asynchronous Serial Interface on

9600 Baud.

5th Managing PBX Interface Based on PBX Message Protocol, and

6th Keeping transaction log for obtaining data for diagnosis and debiting.

The RPU device software controls one serial interface for the PBX call processor 24 and serial interfaces for each of the CCUs 18 in the base station.

The hardware of the RPU is implemented in a Motorola multifunction computer of type 68000. This computer has a 1 Mbyte RAM memory and a non-erasable 10 Mbytes disk space. For input / output, there is a system console as well as a device with 8 asynchronous serial data interfaces.

As shown in FIG. 5, RPU software simulates a system including a scheduling module 40 (scheduler), BCC interface modules 41a, 41b, ... 41n, a PBX interface module 42, a console module 43, a logger module 44, a message processing module 45 (MPM) and a database module 46.

All modules except the database module 46 are controlled from the scheduling module 40. They communicate with each other through a system of mailboxes. Database module 46 is based on a collection of subroutines for accessing information in the database.

The scheduling module 40 creates mainline code for RPU software. It is responsible for planning and activating all other modules and for maintaining event timers and mailboxes that enable intra- and interprocess communication.

The BCC interface modules 41a ,. .. 41n controls a serial, asynchronous interface and a link level protocol. They also monitor the state of communication with the CCUs 18.

PBX interface 42 controls a serial, asynchronous interface to the PBX call processor 24.

The console module 43 creates system operator interface which enables system status queries and modifications and message between RPU 20 and the rest of the system.

The logger module 44 creates raw transaction information for diagnosis and system analysis.

The message processing module 46 processes all the received RCC, BCC and PBX messages. It performs all subscriber call establishment and degradation that is not performed by the PBX 15, and it allocates the radio channel CAN 175353 B1. It also includes a background task that monitors the state of CCUs 18.

The database module 46 creates a consistent interface to all the data structures required for call processing. It includes a frequency allocation task for allocating the radio channels.

The RPU database includes a structure that describes the system configuration, including information about all subscribers and the state of all radio10 channels. These structures are described below:

The RPU database contains a BCC data structure for each CCU device 18 in the system.

A table of subscriber identification numbers (SIN table) contains an ordered list of all approved subscribers. The list is arranged to make subscriber validation easier. The SIN table has one input for each subscriber in the system.

RPU software performs part of the processing of subscriber calls. This processing takes place in the message processing module. The call processing is done by means of message exchanges between MPM 45, PBX module 42 and all BCC modules 41.

Initiating a telephone call from a subscriber station.

This section provides a brief description of the normal dialing procedure for a telephone call from a subscriber. The calling subscriber removes the handset, dials a valid telephone number (the destination number) and press the send key or wait a moment. The calling subscriber station sends a CALL REQUEST message over the control channel to the base station. The BCP modules 41 of the RPU receive the RADIO REQUEST message and forward it to the MPM module 45. This module performs simple validation of the rotated digits and sends a RADIO

REQUEST message for PBX module 42 leading with /.

forwarding to the PBX control processor 24. The PBX call processor 24 validates these digits and sends a PLACE CALL message back to the RPU 20. The MPM module 45 assigns one calling interval to the calling station. MPM module 45 generates a CHANGE CHANNEL order to CCU 18, which contains the voice range assigned to the calling subscriber station. The MPM module 45 generates a CALL CONNECT order to the calling subscriber station which assigns the voice frequency and interval to the calling subscriber station. MPM module 45 generates an ALLOCATE message to the PBX call processor 24, whereby it is ordered to allocate a message channel. At this point, the calling subscriber station is ready. It is now awaiting a connection to the called station through the PBX switch matrix 25. The called station can be either a subscriber station or a telephone to which one is connected over a connection line 14 from a switchboard.

Receiving calls at a subscriber station.

This section provides a brief description of how an incoming call to a subscriber station is processed. The PBX call processor 24 determines whether a telephone call is intended for a subscriber station. The PBX call processor 24 generates an INCOMING CALL message containing information about the nature of the incoming call, namely whether that call arrives from an external connection line 14 or from another subscriber station. The PBX module 42 of the RPU receives the PBX message from the PBX call processor 24 and passes it to the MPM module 45. If the call arrives from another subscriber station, the MPM module 45 sets the subscriber-to-subscriber index for both the calling station and the called station and order the people involved /.

prepared CCUs 18 to switch to internal mode. MPM module 45 generates a PAGE message for the subscriber station specified in the INCOMING CALL message. The station in question answers with a CALL

ACCEPT message. MPM module 45 responds to the CALL ACCEPT message by generating a CHANGE CHANNEL message to the appropriate CCU unit 18 and a CALL CONNECT message to the appropriate subscriber station. Next, MPM module 45 generates an ALLOCATE message 10 to the PBX call processor 24, thereby causing the PBX switch matrix 25 to make the final connection for the oncoming call.

Recovery after failure.

This section provides a brief description of how the RPU 20 responds to the channel outcome during an ongoing conversation. The CCU unit 18 in charge of the voice channel can see that the channel synchronization is lost when there is a failure. The CCU 18 produces a NO-SYNC message. The BCC module 41 receives this message and passes it to the MPM module 45 which then sends an ON-HOOK message to the PBX call processor 24 and brings the subscriber station into idle mode and the channel in 0N-H00K mode.

Processing an arriving BCC message.

A BCC message is transmitted from CCU device 18 to RPU device 20 through a 9600 Baud asynchronous interface. The BCC module 41 that deals with that person

CCU interface, reads the message and checks the link-level information bits to check the integrity of the oncoming message. If the BCC module 41 finds that the message is acceptable, appropriate acknowledgment will be returned to the broadcasting CCU unit

18th Otherwise, a message will be sent to renew the trial or a negative acknowledgment. The BCC module 41 now sends the message to the MPM module 45. This message is placed in the message processing mailbox 48 among the mailboxes found at the scheduling module 40 (cf. Fig. 6).

If there is no more input from the CCU 18, and if the BCC mailbox 49 containing outgoing messages to the CCU is empty, the BCC module 41 interrupts and the control switches to the scheduling module 40.

The scheduling module 40 activates the next module in the round (robin schedule) and this module runs until a shutdown occurs, after which the scheduling module again activates another module etc. At some later time the scheduling module activates the MPM

45.

The MPM module 45 then reads the BCC message along with any other message that may have been queued in the mailbox 48. The BCC message is identified and processed. Such processing may include changes to the database and the generation of new messages. FIG. 6 shows the data path of an incoming message.

Generating an outgoing BCC message.

FIG. 6 also shows the data path of an outgoing BCC message. An outgoing BCC message is generated by the MPM module 45 when some particular event occurs. The message is built up within the MPM module 45 and transmitted to the BCC module 41 serving the called CCU unit 18. After this message and any other necessary messages have been sent and if there are no more messages in MPM's mailbox 48, the MPM module makes a shutdown and the control switches back to the scheduling module.

The BCC module reads the message from the associated mailbox 49 and adds the outgoing message the appropriate link-level bits, upon which the message is sent in serial form to the CCU unit 18.

Processing of RCC messages.

An incoming RCC message is handled exactly like an incoming BCC message since the RCC message is a form of BCC message. Thus, an outgoing RCC message is generated which is sent in the same way as an outgoing BCC message.

Processing of an incoming PBX message.

A PBX message is received from the PBX call 15 processor 24. This message is fed to the RPU 20 through an 9600 Baud asynchronous interface. The FIG. 7, the RPU PBX module 42 takes over the PBX message and sends it to the MPM mailbox 48. When there are no more arriving characters and when the PBX mailbox 50 containing outgoing PBX messages is empty, the RPU makes the PBX module 42 and the control is transferred to the planning module 40.

MPM module 45 acquires the PBX message together with other messages that may have been queued in the associated mailbox 48. The PBX message is processed based on the message type and the current subscriber status specified in the message. The processing may include changes in the database, changes in subscriber status and the generation of new messages. FIG. 7 shows the data path of the arriving PBX message.

Generating an outbound PBX message.

Referring to FIG. 7. An outgoing PBX message is generated by the MPM module 45 in response to an event. The message is constructed in the MPM module 45 and delivered to the PBX module 42. After sending this message and any other necessary message, and if there are no more messages in the MPM mailbox 48, the MPM module 45 interrupts and the control is transferred to the planning module 40.

The scheduling module 40 continues to activate other modules according to the above-mentioned round robin principle until the RPU PBX module 42 is activated.

The RPU PBX module 42 receives the PBX message from the mailbox 50 and sends the message in serial form to the PBX call processor 24.

Generating Logger Messages.

At relevant points in each module of the RPU program20, a message containing relevant information is sent to the logger module 44. This information is timestamped and sent to a file. FIG. 8 shows the paths for logger data.

Console input / output module.

The entry portion of the console module 43 provides for order creation, recognition and order validation. Valid console orders are able to query and update the RPU database and send messages to the RPU modules. The output signal from console display orders is output directly to console gate circuits.

Grading Module.

The planning module 40 is a special system module which is responsible for controlling all others

RPU modules. The main responsibility of the planning module 40 is to select the next module to operate and create inter- and intra-module communication.

While the various RPU modules can be perceived as separate modules, all of these modules are in fact a single use of a Regulus operating system. It is the scheduling module 40 that performs the aforementioned round-robin distribution to the other RPU modules. The scheduling module 40 controls the stack for each of the RPUpseudo modules by allocating a given portion of the stack15 space to each of the pseudo modules at the startup time. Immediately before each module has to be started, the scheduling module 40 brings a stack arrow to change point to the appropriate stack address for that module. FIG. 9 shows a storage card for the RPU module 20.

Each RPU module runs until it is blocked.

When one module is blocked, the control is transferred back to the scheduling module, which then lets another module run. A module can be blocked in various ways by inducing GETEVENT (), which forces the module to block until an event is in progress, or by generating WAIT (), which blocks the module for a given number of seconds, or by producing BLOCK ( ), thereby blocking over one lap in said round-robin scheduling loop.

Another main function of scheduling module 40 is inter-module communication between the modules. Said mailboxes are used as the means of sending or receiving messages to or from other modules. Each module can check if there is shipping in its mailbox using the MAILREAD () signal. Similarly, a module can broadcast / · to another module using the MAILSEND () signal. The scheduling module keeps a separate mailbox for each of the modules included in the scheduling loop. When a module sends a message to another module, the message is copied in the addressed mailbox. When it is later the mailbox to run, the scheduling module examines its mailbox to determine if there is a message in it. If so, the scheduling module 40 produces an event of the type MAIL which forces the module to be unblocked if it had been blocked by GETEVENT (), allowing the module to run.

The scheduling module also keeps a list of the events for each module in the loop. These events can be mail (mail) or time events. Mail events are generated when the scheduling module finds that there are messages for the current module. A module can set a time event on the list by calling PUTEVENTO with the number of seconds to wait before an event is created. The scheduling module 40 checks the module's event list at each turn in the loop to search for timeouts. When timeouts are detected, the appropriate module is run and the event is returned to the module through the GETEVENT () signal.

The scheduling module 40 contains routines used to initialize RS-232 interfaces between CCU unit 18 and RPU unit 20 and between PBX unit 15 and RPU unit 20. These routines take exclusive software control over RS-232. interface, interrupts the usual processing of the control sequences of the Regulus operating system. Other routines are used to clear the I / O buffer circuits and to load and read final inputs and outputs. The scheduling module 40 also keeps an eye on system timing for all RPU modules.

/·

BCC interface module.

Each BCC module 41 creates an interface between a CCU device 18 and the other software modules in the RPU device

20th The messages between CCU unit 18 and RPU unit 20 consist of variable length binary data which is transmitted over an asynchronous communication connection.

<td></td><td>BCC module</td><td> 41</td><td>have</td><td>responsible for</td><td colspan="3">message integrity</td>

<td></td><td colspan="2">over the connection</td><td>, der</td><td>include</td><td colspan="2">fault detection.</td><td>SEC-</td>

<td> 10</td><td>friend's formation</td><td>for</td><td colspan="2">the messages and</td><td>recognition</td><td>of</td><td>with-</td>

<td></td><td>delelserne.</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td></td><td colspan="4">The hardware interface between</td><td>CCU unit</td><td> 18</td><td>and</td>

<td></td><td>RPU device</td><td> 20</td><td>is a</td><td colspan="2">asynchronous interface RS-232</td><td>on</td><td> 9600</td>

<td></td><td>Baud.</td><td></td><td></td><td></td><td></td><td></td><td></td>

<td> 15</td><td>input</td><td>to</td><td>this</td><td>module 41</td><td colspan="3">includes message</td>

view from the CCU device or from other RPU software modules. Messages are emitted from this module either to the CCU via the RS-232 interface or to other RPU software modules via the appropriate mailbox.

The purpose of module 41 is to process the message traffic between the RPU 20 and the CCU

18th This module 41 continuously checks the messages arriving from the CCU 18 and directs them to the respective RPU software module. Like 25, this module continuously checks for messages from other RPU software modules to a CCU device

18th An alternate bit protocol is used to limit pending messages, ie. unrecognized messages, to one in each direction. Sequence and recognition bits serve as the necessary route control to perform this function. The protocol is described in more detail in the following sections.

In the following statement, one entity is capable of processing messages by us or us, while the other is termed by them or them. The ProtoDK 175353 B1 coli can be explained by specifying the measures taken when a message is received. There are only four basic measures that depend on two conditions. These conditions are determined by comparing the sequence and recognition bits in the received message with the expected bits.

In an oncoming message, the ACK bit is as expected if it is the same as the SEQ bit in our last broadcast message. Likewise, the SEQ bit is as expected if it deviates from the SEQ bit in the last received message. In other words, the expected conditions are that an incoming message recognizes our last message, and we also expect each incoming message to be a new message.

The precautions taken when receiving a message are now summarized for four combinations based on the above conditions:

First ACK as expected; SEQ as expected. Our last sent message is marked as acknowledged, which allows us to send a new message. The newly arrived message is processed, ie. is acknowledged in the next message we send.

2nd ACK as expected; SEQ not as expected. Our last sent message is marked as acknowledged, which allows us to send a new message. The newly arrived message is discarded, ie. is not recognized.

Third ACK not as expected; SEQ as expected. If we have sent a message that has not yet been acknowledged, we will resend it. If we do not have such a message, something is wrong at the destination and we need to reset (Reset) as described below. The newly arrived message is being processed.

ACK not as expected; SEQ not as expected. Our last message was not received at the destination. We will reissue the message and discard the newly arrived message.

The reset bit is used to reset the SEQ and ACK bits. When we receive a reset bit message, the message must be accepted as a new message regardless of its SEQ bit and this message should be acknowledged.

The ACK bit on the received message reflects the SEQ bit in our last message that they received from us. We should toggle this bit before sending the next message. For example, if we receive a message whose ACK / SEQ digit is 4 (Reset = 1, ACK = O,

SEQ = 0), the ACK / SEQ digit in the answer should be 1 (Reset = 0,

ACK = 0, SEQ = 1). Both parties can reset when they believe the protocol has expired.

When we receive a message from them and have no new message waiting or a default response that does not arrive immediately, we will acknowledge the message by sending a special ACK message. The ACK bit recognizes the received message, but the SEQ bit does not change from the last message we sent. This causes the counterparty to process the acknowledgment and cash in on the newly arrived message. The content of this message is a zero message. In any case, since this message is discarded, the content of the message is irrelevant.

PBX interface module.

The PBX module 42 interfaces between the UTX-250 PBX call processor 24 and the other software modules of the RPU 20. The messages between the two devices are ASCII standard. The ASCII character is defined here as 7 or 8 bit ASCII. Both the PBX call processor 24 and the RPU 20 must be able to accept characters down odd or odd parity or without parity. The text of the messages consists of variable length chains or printable characters.

The hardware interface between the PBX call processor 24 and the RPU 20 is an asynchronous interface RS-232 of 9600 Baud.

starting from or to

Input to the PBX module 42 includes messages received from the PBX call processor 24 or from other RPU software modules. Announces this module to the PBX call processor 24 other RPU software modules via that mailbox.

The purpose of the PBX module 42 is to process the message traffic between the RPU 20 and the PBX call processor 24. This module continuously checks the received messages from the PBX call processor 24 and directs them to the respective RPU software module. Also, the module continuously checks for messages from other RPU software modules to which the PBX call processor 24 must connect.

Each character received by the PBX call processor 24 is checked for similarity to the character&gt; indicating the beginning of a message or a sled return sign indicating the end of a message. This module is capable of handling message traffic in full duplex.

Console Module.

The console module 43 is the window that gives the operator access to the current status of the RPU 20. The console provides the ability to continuously display subscriber and radio channel status information, change the connection and channel status, and send messages to the PBX 15 and CCUs 18. The console processes the incoming power from the terminal and executes the desired order.

Console module 43 creates interface to the base station operator terminal. Console module 43 processes input from the terminal and executes the order. The data is retrieved from and loaded into the database, the information is displayed on the terminal screen and messages are sent to other modules. This module's interfaces include:

1) Characters are entered through the operator's keyboard.

2) Characters are output to the operator screen.

3) Data is retrieved from or loaded into the database.

4) Messages are sent to PBX, BCC, and message processing modules.

A set of analysis routines loads characters from the operator keyboard. Data input call is displayed at the beginning of each order line, the data is inserted into buffer circles, the editing characters are processed, the input data is displayed on the screen and the data is delimited by labels. When the analytics routine is provided with a set of data structures that describe all possible orders and valid labels within each order, the analytics routine is able to recognize the data entered, respond to queries, and display data entry keywords. Each brand is checked to see if it is the type of data you expect. The keywords are compared to a list of acceptable inputs, and numbers (numbers / numbers) are converted to integers. When the loaded order line is complete, additional checks are made. The numbers are checked to see if they are within a given range and for some orders the system status is checked before the order is executed.

The orders are divided into three groups: 1) order to display information from the database, 2) order to change the database and 3) order to send messages. Subscriber, connection, CCU and channel status information can be displayed. All display orders require retrieval of information from the database and formatting of data to the operator screen. The modification orders include the ability to force a subscriber connection to a given channel and the ability to unlock and lock channels. These modification orders are used to control the frequency allocation algorithm. All modification orders load into the database.

From the console module 43, PBX, BCC and RCC messages can be sent to various other modules in the syste /.

system. A SENDMSG order prompts the operator to provide all necessary information for the message, after which the message is formatted and passed on to the specified module. PBX messages are sent to the RPU PBX module 42 which sends the message out to the PBX call processor 42.

BCC and RCC messages can be sent from the RPU 20 to the CCUs 18 via the BCC modules 41 which add to the outgoing messages the link level protocol bits. Input from CCUs 18 is simulated and message 10 claimants including both the BCC and RCC messages are passed to the MPM module 46.

Logger Module.

Logger module 44 is responsible for logging off

RPU events or messages. The logger module 44 holds the following three disk files: a transaction log with information such as billing information, an error log of faulty messages, and a message log consisting of system alert messages.

The logger module 44 consists of a set of subroutines called from the other RPU modules. Each subroutine is responsible for timestamping the message and loading the message into that disk file. Each subroutine has a global flag that specifies whether to log messages or not. These flags are set and reset with orders from the console.

Message Processing Module (MPM).

MPM module 45 performs high-level call processing functions between the PBX 15 and the subscriber stations. It is responsible for such call processing capabilities as initial searches, voice channel allocation, and call progress toner management for both subscriber and external telephones. The MPM35 module 45 also processes status messages that it receives from the CCUs 18. For example, the MPM module 45 processes channel status information regarding the connection quality or touch-on condition of the subscriber.

MPM module 45 is organized as a state machine where PBX and BBC messages are tokens for the message processing state machine. MPM module 45 processes these tokens by updating the database, providing the required responses, and then transitioning to the next state.

MPM module 45 uses the system mailbox maintained by scheduling module 40 for receiving and transmitting messages to and from other RPU modules. MPM module 45 uses subroutines in the database module to record or update state information in the database.

As mentioned above, the MPM module 45 is organized as a state machine. Tokens that force processing consist of messages or timeouts. MPM module 45 determines the type of token (i.e. timer, RCC message, PBX message, etc.) and the subscriber station or channel to which the token relates. MPM module 45 processes the token by generating the correct message responses and causing the transition to the next state.

The MPM module 45 actually consists of two state tables. The FIG. 10, the RCC state machine is used to process messages from the PBX call processor 24 or RCC messages from a subscriber station. The FIG. 11 channel state machine is used to process messages received from a CCU device 18.

To begin with, all subscribers are in RCC idle mode and all channels in channel idle mode, indicating that no connections or connections have been established.

For a typical outside call to a subscriber, the state changes are as follows. An external call withDK 175353 B1 sharing is received from the PBX call processor 24, which message includes the telephone number of the subscriber station being called. A PAGE message is sent to the subscriber station and the subscriber station state changes to PAGE. When a CALL ACCEPT message is received from the subscriber station, the subscriber station state changes to ACTIVE. At this point, a channel is assigned and the PBX call processor 24, the CCU 18 and the subscriber station are notified of the channel assignment. The channel is placed in a RING SYNC-WAIT mode (Fig. 11) · When the CCU 18 indicates that synchronization has been achieved, the channel mode changes to SYNC RING. When the CCU 18 finally indicates that the subscriber has taken the tube off, the channel switches to SYNC OFFHOOK mode. This SYNC OFFHOOK mode indicates that there is an established voice connection.

A subscriber-to-subscriber call begins with a CALL REQUEST message received from the calling subscriber station. The calling subscriber station switches to DIAL mode, and a RADIO REQUEST message is sent to the PBX call processor 24. The PBX call processor 24 sends a PLACE CALL message to the calling subscriber station and an INCOMING CALL message to the called subscriber station. In response to the PLACE CALL message, a channel is allocated and the PBX call processor 24, CCU 18 and the calling subscriber station are notified of this assignment. The channel state of the calling subscriber switches to OFFHOOK SYNC WAIT until the channel becomes synchronized. When the base station's CCU defects the transmission of the calling subscriber, it generates a SYNC OFFHOOK channel message. The RPU 20 processes this channel message by changing the channel state to the SYNC OFFHOOK mode. An incoming call message to the called subscriber station is processed in the same way / · as an external call message as described above. In addition, the channels involved in the connection are set to internal mode when both subscriber stations are synchronized.

The breakdown begins when one of the parties in the connection hangs up. When an external device is piped, the MPM module 45 receives from the PBX call processor 24 an ONHOOK message. When a subscriber station switches to ONHOOK, the CCU 18 sends a message indicating that the subscriber station has switched to ONHOOK. In either case, the counterparty is notified of degradation, the channel switches to DISCONNECT mode, and the subscriber station switches to the TEARDOWN state. When the CCU 18 indicates that the synchronization has been lost, the channel and the subscriber station return to idle mode.

Background Tasks.

A background task routine is implemented in the MPM20 module 45. The background task initially connects to the CCUs 18 after a cold or hot restart. When the system is operational, the background task monitors the CCU units 18 to keep the database up and assigned an RCC unit.

BCC messages generated by both CCUs 18 and BCCs 41 are received from BCCs 41. Messages are sent to CCUs 18 via BCCs 41. The data is loaded into and read from the database.

Initially, all CCU units 18 are sent

BASEBAND QUERY messages so that the RPU 20 can determine the current state of the system. All information received from baseband event or response messages is stored in the RPU database. When the RPU 20 receives a baseband event message indicating that a

CCU unit 18 is clear and not reset (for example, the CCU unit is not yet fully plugged in), the frequency assigned to the CCU unit 20 is marked as being allocated. The CCU unit 18 is then sent

CHANNEL QUERY messages for updating the database according to the current state of the system. The CCU initialization is complete when each CCU device 18 has either responded to all pending query messages or when the CCU device 18 is found to be disconnected. At that point, each CCU 18, indicating that it was ready and reset (i.e., the CCU has just been plugged in), is assigned a frequency. If no control channel is assigned to a CCU unit 18, the RPU unit 20 attempts to assign the control channel. The first choice is to assign the control channel CCU unit 18 at the first frequency, being that the subscriber first searches for the RCC unit. The next choice relates to any CCU device 18 where the O interval is not in use and the last choice relates to a CCU device

18 with a compound in the O interval. If all already operating CCUs 18 already have a connection in the O interval, one of the connections is terminated in the O interval and the control channel is assigned to this interval.

5 Once the RPU 20 has communicated with all CCUs 18, the status of CCUs 18 is monitored via status messages received from CCUs 18 or BCC modules 41. BCCs 41 continuously monitor the communication 30 path to each CCU Unit 18. A CCU unit 18 is considered to be out of service when a baseband event message is received indicating that the CCU unit is not ready. At this point, the CCU 18 is marked as not ready in the database. Furthermore, all connections are broken and all channels are returned

7Ö to default mode and the frequency assigned to the CCU 18 is dislocated. If the CCU 18 contains the control channel, a new control channel is assigned.

When a baseband event message 5 is received indicating that a CCU unit 18 is ready and reset, the CCU unit is assigned a frequency. If no control channel is assigned to a CCU 18 at this time, the remaining CCU unit's O interval control channel is assigned.

If a baseband event message is received indicating that a CCU device 18 has lost communication with the RPU 20, the CCU 18 sends CHANNEL QUERY messages (one for each of the four channels) to update the RPU the database with the current state of each of the CCU channels. When a response is received to each CHANNEL QUERY message, the current channel state and connection information is updated in the database. If a channel is in SYNC WAIT mode, it is assumed that the subscriber is no longer involved in the connection and this connection is disconnected.

Initially, the RPU 20 interrogates the CCUs 18 about their initial state. The CCUs 18 also issue event messages each time they are turned on or when they change state. With the 25 sharing exchange, the RPU database keeps up to date with the current state of the system.

Database Module.

The database module 46 contains the necessary data interface database routines to access the database. They create a concise, single-threaded interface to the database for any module that requires access to the information contained in the database. Most of the access routines are linked to the SIN table and the BCC-ta35 bell. The access to all fields in these tables is created by the access routines.

/...

The database module is also responsible for initializing the database at start-up. All significant fields are initialized to appropriate values ​​using the database module initialization part.

The database module creates the following routines:

1) TTY initialization procedures,

2) a subscriber search binary search routine in the SIN table,

3) frequency-to-CCU10 mapping routines and tables,

4) managing diagnostic display information; and

5) frequency allocation.

Database module 46 is a collection of routines that ensure the controlled access of other modules to the database. By managing all approaches through the database routines, the database is kept essentially hidden from external modules. This allows you to change the database without requiring modifications in the other modules. When the database is changed, only the interface routine for the changed part of the data20 database needs to be changed.

Frekvensallokeringsopgave.

The frequency allocation task performed by

The purpose of the RPU 20 is to select an appropriate frequency and range for a subscriber station requesting a voice channel. The selection algorithm takes into account the type of call (internal or external) and the modulation level (16-level or 4-level).

Although the frequency allocation task is functionally independent of the database module 46, it is closely related to the data structures in the database. As a result, this function is described separately from the database module, although it is technically a routine in the database module 46.

The frequency allocation task is used by the MPM unit during call setup. It makes strong use of the data structures in the database module.

All frequency allocation requests fall into one or the other of two categories. The first concerns external sources and the second internal destinations. The Internal Destinations category covers the arriving portion (destination) of an internal call. The External Sources category covers all other cases that include external calls, whether for incoming or outgoing calls, or the origin of an internal call.

Input for the frequency allocation task consists of an index to SIN table for the subscriber station requesting a channel and the index to SIN table for the calling subscriber station. The calling subscriber station index is only valid when the channel is set up for an internal destination call. In all other cases, the calling subscriber station index is a predetermined invalid index defined as DB NULL. These indexes provide access to all the information needed to allocate a suitable channel, ie. frequency and range.

The frequency allocation routine returns TRUE if it succeeds in allocating a combination of frequency and range; otherwise it returns FALSE. The selected frequency and range, if allocated, are loaded into the SIN table of the subscriber station requesting frequency assignment.

Each frequency is divided into four TDM intervals. The RPU database records how many intervals are available in each position. When a request for allocation falls within the category of external sources, an interval is selected in the interval position corresponding to the largest number of free positions. When an interval position is selected, the first available frequency is selected in that interval. When a call falls into this category, it actually has none

/...

whichever interval is chosen, but this technique tends to distribute system load evenly across all intervals, and - which is more important - increases the likelihood of optimal interval allocation for both parties in an internal call. This is true because system timing calculations have shown that the optimal range assignment for a subscriber-to-subscriber call exists when the base station transmitting interval for each subscriber is in the same range at different frequencies. By assigning the calling party to a subscriber-to-subscriber call the most accessible interval position, over time it is more likely that the called subscriber station will be able to assign the same interval position at a different frequency. For example, if position # 2 is the most accessible position, it is selected. When the called subscriber station allocation request is processed, it is more likely that another interval in position no. 2 is available and can be selected, resulting in optimal range-to-range allocation.

When a request for allocation falls within the internal destination category, the range to be assigned in a selection table is selected. The selection box contains lists from the most to the least desirable range position assignment for the called subscriber. This scheme is based on the interval assignment of the calling subscriber. So far, the type of modulation has not been mentioned. This is because the basic rules for allocation do not change when selecting 4 and 16 intervals, except for one important exception, namely that for a 4-type compound, only O-interval or 2-interval can be allocated. . Due to this exception and the fact that two subscribers could be tuned to different moDK 175353 B1 types, a total of four unique selection tables are required to cover all possible call combinations. These tables are listed below.

Table 6

origin range

First choice 2. choice 3 choice 4. choice interval 0 interval 1 interval 2 interval 3 rating—&gt;

(1)

2

2 3

0

2 1 (2a) (2b) (3) preferred range selection table for internal calls from 16-origin to 16-destination.

It should be noted that in each table there is a 20 associated rating indicating the desirability of a given interval. The most desired range will have a rating of 1, and lesser desired ranges of ratings of 2, 3, etc. If two or more bars in a table have the same degree of desirability, they are assigned the same number followed by letters a, b, c, etc. , so that three columns where rating is indicated by 2a, 2b, 2c, etc. have the same degree of desirability in any order.

origin range

First choice

Table 7

2nd selection choice interval O interval 2 rating—&gt;

12

3 0 (1a) (1b) (2a) preferred range selection table for from 4-origin to 16-destination.

4th choice (2b) internal calls

Table 8

Origin interval interval 0 20 interval 1 interval 2 interval 3 rating—&gt;

First choice

2nd choice (1) (2) preferred range selection table for from 16-origin to 4-destination.

internal calls

origin range

Table 9

First choice 2. choice interval 0 interval 2 rating—&gt;

2

0 (1) (2) preferred range selection table for internal calls from 4-origin to 4-destination.

The frequency allocation task has two inputs, which provide access to crucial information, for appropriate frequency and interval selection.

The first input is the SIN table index for the subscriber station requesting a channel. With this index, the frequency allocation can determine the default modulation type of the calling subscriber. It also tells the routine where to place the results of its selection algorithms, ie. frequency number and range number.

The second input to the frequency allocation task specifies the category of frequency / range request. The value of the second input is either an index to SIN table or the invalid value DB NULL defined above. If a valid index is received, the frequency allocation request is identified as the destination part of a subscriber-to-subscriber call, and the selection tables should be used. If DB NULL is received, the request is considered to fall within the external source category and the most available range position algorithm is used.

The frequency allocation task returns to

TRUE if it succeeds in allocating a combination of freDK 175353 B1 sequence and range? otherwise it will return to FALSE. This also carried an undesirable side effect. If the allocation is successful, the baseband index and interval in SIN table fields are filled in for the calling subscriber.

The frequency allocation algorithm can be broken up into two steps. The first step, called the classification step, determines the category of the allocation request. The second step, called the selection step, finds and allocates a combination of frequency and range using the appropriate algorithm determined by the allocation request category.

The classification step first determines whether an automatic frequency selection should take place. If the calling subscriber has been switched in manual mode, the specified values ​​for manual modulation level, manual frequency and manual interval will determine the frequency / range modulation to be allocated. If the specified frequency and range are available, they are assigned to the calling subscriber. If the specified frequency and range are not available, the routine expires and returns to FALSE value. If the calling subscriber has been switched to automatic mode, further classification is required.

After it has been determined that automatic selection must be made, the frequency allocation algorithm determines the request category. These request categories are divided as follows: External-in when a destination subscriber is called from an external telephone;

External-out when a calling subscriber station calls an external telephone? Internal-out when one calling subscriber station calls another subscriber station; Internal-in when a destination subscriber is called from another subscriber station. If the request is for exter35 nal-in, external-out or internal-out, an interval position is selected by a search for the most available position. Once the position has been selected, all frequencies are searched sequentially until a desired interval is found in the desired position (or a few adjacent intervals in the case of a 4-level request). At this point, the routine inserts the appropriate values ​​into the SIN table, after which the routine ends and returns to TRUE. If the request falls into the last category (internal-in), further information is required.

When it comes to an internal-type request, two more bits of information are required. Interval assignment and the modulation type (4 or 16 levels) of the calling subscriber must be identified. Once this has been done, the appropriate selection table is determined based on the modulation type of the calling subscriber station and the destination station. After selecting this table, the calling assignment interval allocation is used to determine the appropriate row to be used in the selection table. Each sequential link in the selected row contains interval allocation with equal or lower degree of desirability. The list is reviewed until one is available. interval, starting from the most desirable position and continuing until all interval positions have been reviewed. For each interval position (or pairs of intervals for 4-level connections), the individual frequencies are examined sequentially until the respective interval or interval pair is found. The resulting frequency and range values ​​are not entered in the SIN table in question and the routine expires and returns to TRUE.

A slot count array is kept up to date with the number of intervals available for each interval position.

These count values ​​are kept in the database module and the frequency allocation task is assigned to these values.

SIN table contains relevant information about each of the approved subscribers in the system. There are the following accesses to the SIN table.

Modulation Level

The modulation level of the subscriber station requesting a frequency is deduced from the table together with the modulation level of the calling subscriber during an internal type connection establishment.

(reading)

Interval number (reading):

Interval number (input):

In a connection type of internal type, the interval assignment for the calling subscriber must be found.

The interval allocation for the subscriber requesting a channel is entered here.

baseband index (load): The frequency assignment for the subscriber requesting a channel is entered here.

The frequency assignment routines search uses BCCtabel for an available combination of frequency and range. The following BCC table accesses are created:

KanaltiIstand

The condition of a channel is examined for the purpose of determining availability.

(reading)

Channel status (reading):

Channel status is checked to check that the specified channel is a voice channel.

Channel Mode (Load):

Channel management (loading):

SIN index (input):

Channel mode changes when the specified channel is selected for assignment.

The modulation type of the requesting subscriber is loaded into channel control bytes.

Establishes a connection from the selected channel to the requesting subscriber.

The frequency allocation routines have direct access to the database. This is due to speed and efficiency. Whenever possible, the database interface routines are used to create the frequency allocation routines access to the database.

Subscriber Interface Unit (STU).

The subscriber interface unit is intended to convert the analog signal arriving from a conventional telephone set into a 64 kbps digital PKM encoded signal. The FIG. 12, the STU unit shown comprises a subscriber line interface circuit 53 (SLIC) which directly connects over wires 37 to a type 500 keypad telephone. The SLIC circuit 53 creates appropriate values ​​for voltage and impedance of the telephone apparatus. The SLIC circuit 53 also sends a ringing tone to the device and defects when the tube is mounted or removed. The output signals from SLIC circuit 53 over line 54 are transmitted and received voice frequency analog signals. These signals are converted to PKM samples in PKM codec 55. PKM codec 55 uses a compression algorithm with compression parameter u = 255 to digitize the speech signals to 8-bit samples at 8 kHz.

PKM codec 55 works full duplex. The line 56 transmits the digitized signal to a mode select multiplexer 57. The mode of operation of the multiplexer is determined by the subscriber station controller SCU 58, which communicates with the multiplexer 57 through transmit / receive FIFO register 59. The SCU circuit 58 generally includes a type 803 microcontroller. The SCU circuit is coupled to CCU 29 through an RS-232 interface circuit 60 and controls the operation of the SLIC circuit 53.

The subscriber interface circuit operates in three separate modes, the first and most common being voice mode, where the speech samples from PKM codec 55 through multiplexer 57 and drive circuit 61 are transferred to speech codec / decoder unit 28, where they are further processed to reduce the bit rate from 64 kbps to 14, 6 kbps, after which the signal is transmitted to the base station.

The second mode of operation is data mode in which the 64 kbps stream to and from the VCU 28 does not contain telecommunication information. The information for the base station is a reformatted data stream from an external data source with amount of data up to the channel's data transmission capacity of 14.6 kbps. The STU single bed includes an RS-232 data port 62 for connecting to a data device, e.g., data terminal, through a line 63 using a standard 9600 Baud asynchronous RS-232 interface. The STU includes a UART and a timer circuit 64 for synchronizing the data from the RS-232 data port 62. The VCU 38 transforms the synchronized data into packets so that they can be transmitted within the channel limit of 14.6 kbps . In this mode, data transmission is at full duplex.

The third mode for the STU device is call setup. In this mode, no data is transmitted from the STU unit 27 to the VCU unit 28 through the multiplexer 57, but this multiplexer is connected to a ringback tone generator DK 175353 B1 circuit 65. This circuit performs digital synthesis of the tones used in the call procedure, e.g. busy tone and error tone. During the call setup, the user entered is defective

DTMF signals by means of a DTFM detector 66 and processed by SCU circuit 58 for the location of the call. The tone generator 65 sends appropriate tone signals to the user's headset. A ring generator 67 is connected to the SLIC circuit 53. A timing generator

68 generates timing signals for PKM codec 55, driver circuit 61 and ringback tone generator 65. When the call placement is completed, the STU switches to voice or data mode for communications with the base station.

The STU unit must also provide suppression of unwanted echo signals from remote connections. The delay for voice signals from the base station to the subscriber station and back is more than 100 ms. A reflected signal due to incorrect impedance matching at one end or the other gives an annoying echo signal. The problem is solved in the base station using an echo suppression system in the PBX function. The STU must provide echo suppression with the subscriber. At this suppression, an attenuation of the echo25 signal of at least 40 dB is expected. However, the delay of the echo signal to be suppressed is very short since the reflection occurs between the SLIC circuit 53 of the subscriber interface unit and the local telephone apparatus. Normally this distance will be only a few meters and the delay is negligible.

Microprocessor Controller 8031 ​​of the SCU circuit performs the functions of the RPU 20 and the PBX call processor 24 of the base station. It communicates with the base station RPU 20 through the messages sent over the radio control channel and it controls

/.

all the individual functions of the subscriber interface unit. The SCTU also communicates with the subscriber station CCU 29 through the baseband control channel. The RS-232 interface for the CCU 29 operates on the 9600

Baud and used to transmit control information between CCU 29 and STU 27 of the subscriber station.

Voice codes / decoder unit (VCU).

Four full-duplex RELP speech compression systems have been implemented in the VCU. There is identical VCU design for both the base station and the subscriber stations. At the subscriber station, only one quarter of the global functionality is used, namely only one of the four channels. Interface to the STU unit 27 of the subscriber station is identical to the interface used by the four PBX channels of the base station VCU unit 17. The VCU 17, 28 uses a completely digital system to implement the RELP speech algorithm, cf. U.S. Patent Application U.S. Patent Application No. 667,446 of November 2, 1984 regarding RELP Vocoder Implemented in Digital Signal Processors. Alternatively, a sub-band codec may be used. The processed data is fed to CCU unit 18, 29 over a common parallel bus interface controlled by CCU25 software. The CCU 18, 29 transmits to the VCU

17, 28 control signals determining operating mode and configuration of the VCU 18, 29. Operating mode, a functional description and considerations regarding the implementation associated with the VCU 17, 28 are described in more detail below.

FIG. Figure 13 shows the interfaces between PBX 15 and VCU

17, while FIG. 14 shows the interfaces between STU 27 and VCU 28. The STU 27 interfaces are a subset of the PBX 15 interfaces in that the STU 27 provides only one full-scale speech operation. The timing relationships between

The PBX and STU interfaces are identical, cf. 15. The symbols used in FIG. 15 are specified in Table 10.

Table 10

<td></td><td>Symbol</td><td>parameter</td><td>Mine</td><td>Typ</td><td>Max</td><td>Unit</td>

<td></td><td>Two</td><td>PBX breaks width</td><td> -</td><td> 125</td><td> -</td><td>PS</td>

<td></td><td>TWL</td><td>clock pulse width</td><td> 1,8</td><td> 2,0</td><td> 2,2</td><td>PS</td>

<td> 10</td><td>tw2</td><td>gate 0 inactive width</td><td> -</td><td> 93,75</td><td> -</td><td>PS</td>

<td></td><td>TW3</td><td>gate 0 inactive - gate 1 width</td><td> 5,9</td><td> 7,8</td><td> 9, 7</td><td>PS</td>

<td></td><td>tw4</td><td>gate 1 inactive - gate 0 width</td><td> 52, 8</td><td> 54, 7</td><td> 56, 6</td><td>ps</td>

<td> 15</td><td>TDO</td><td>start pulse - clock 0 delay</td><td> 0</td><td> 250</td><td> -800</td><td>ns</td>

<td></td><td>tdl</td><td>start pulse - clock 1 delay</td><td> 0</td><td> 250</td><td> -800</td><td>ns</td>

<td></td><td>td2</td><td>clock 0 - gate 0</td><td></td><td></td><td></td><td></td>

<td> 20</td><td></td><td>flank delay</td><td> 100</td><td> 1000</td><td> 2000</td><td>ns</td>

<td></td><td>td3</td><td>clock 1 - gate 1 flank delay</td><td> 100</td><td> 1000</td><td> 2000</td><td>ns</td>

<td></td><td>Tso</td><td>input data setup time</td><td> 20</td><td> 1500</td><td> -</td><td>ns</td>

<td></td><td>TSL</td><td>output data setup</td><td></td><td></td><td></td><td></td>

<td> 25</td><td></td><td>hour</td><td> 500</td><td> 1800</td><td> -</td><td>ns</td>

<td></td><td>Tho</td><td>output data hold time</td><td> 500</td><td> 2200</td><td> -</td><td>ns</td>

Referring now to FIG. 13. PBX SDATO, 1st, 2nd and

Third lines 70, 71, 72 and 73 transmit data signals from

PBX 15 to VCU 17 in the base station. In the subscriber station, the data signal is transmitted over STU SDATO, line 74 from STU 27 to VCU 28 (Fig. 14). Serial 8-bit V-255 compressed data is applied to voice codec during the active part of PBX / STU GATEO or PBX GATE1 ... 3 with

5 a frequency of 256 kHz. The data is input into VCU 17, 28 on the leading edge of the pulses in the 256 kHz clock signal.

/·.

VCU SDATO 1st, 2nd and 3rd lines 75, 76, 77, 78 transmit data signals from VCU to PBX 15 in the base station. At the subscriber station, the VCU SDATO line 29 transfers data from VCU 28 to STU 27. Serial 8-bit u-255 compressed data is transmitted from voice codec to PBX 15 or STU 27 during the active high portion of PBX / STU GATEO or PBX GATE1 ... 3 at the frequency of 256 kHz. The data is read from VCU 17, 28 on the leading edge of the pulses in the 256 kHz clock signal.

PBX GATEO, 1, 2 and 3 lines 80, 81, 82 and 83 transmit gate signals from PBX 15 to VCU 17 at the base station. The STU GATEO line 84 transmits a street signal from STU 27 to VCU 28 at the subscriber station. The gate signal is an active high signal used to allow transmission of PBX / STU SDATO, PBX SDAT1 ... 3 and VCU SDATO ... 3. This signal is active for eight successive clock periods every 125 µε.

PBX CLKO, 1, 2 and 3 lines 85, 86, 87, 88 transmit 256 kHz clock signals from PBX 15 to VCU 17 at the base station. The STU CLKO line 89 transmits a 256 kHz clock signal from STU 27 to VCU 28 at the subscriber station. A 256 kHz clock signal is used to input PBX / STU SDATO and PBX SDAT1. . . 3 signals in VCU 17, 28 and VCU SDATO ... 3 signal in PBX 15 or STU 27.

However, these clock signals are not synchronized with any of the clock signals produced within VCU 17, 28, CCU 18, 29 or modem 19, 30.

In the base station, the PBX-VCU interface converts four channels of synchronous 64 kbps serial data into 8-bit parallel data, which is then made available to the four transmit voice codecs 16 at a sampling frequency of 8 kHz. In the subscriber station, there is only one channel (channel 0) that is converted by the STU-VCU interface. The necessary clock signals and gates are created by PBX 15 and STU 27.

The PBX-VCU and STU-VCU interfaces also perform the complementary function of the receiving voice codecs. In the base station, 8-bit parallel data received from the four codecs is converted to four serial, synchronous 64 kbps channels for transmission back to PBX 15. In the subscriber station, only a single voice channel is sent back to STU 27.

FIG. 16 shows the hardware interfaces between the VCUs

17, 28 and CCU 18, 29. FIG. 17 and 18 show the timing constraints between VCU and CCU for the transmit channel and the receive channel, respectively. The Symbols used in 17 and 18 are specified in Table 11 and Table 12.

Table 11

<td>Symbol</td><td>Specification</td><td>Mine</td><td>Max</td><td>Unit</td>

<td>tdl</td><td>voice codec block transfer</td><td></td><td></td><td></td>

<td></td><td>period</td><td> -</td><td> 750</td><td>usee</td>

<td>td2</td><td>TCVC response time</td><td> 1,25</td><td> 15</td><td>ysec</td>

<td>td3</td><td>CCU DMA response time</td><td></td><td> 1,25</td><td>pSEC</td>

<td>td4</td><td>handshake delay</td><td></td><td> 15</td><td>nsec</td>

<td>TD5</td><td>VC block period delay</td><td></td><td> 150</td><td>pSEC</td>

<td>for</td><td>control data hold</td><td></td><td></td><td>nsec</td>

<td>th2</td><td>status data hold</td><td></td><td></td><td>nsec</td>

<td>th3</td><td>TC data hold</td><td></td><td></td><td>nsec</td>

<td>TSL</td><td>control data setup</td><td></td><td></td><td>nsec</td>

<td>ts 2</td><td>status data setup</td><td></td><td></td><td>nsec</td>

<td>TS3</td><td>TC data setup</td><td></td><td></td><td>nsec</td>

<td>TWL</td><td>write width</td><td></td><td></td><td>nsec</td>

<td>tw2</td><td>read width</td><td></td><td></td><td>nsec</td>

<td>TW3</td><td>block request width</td><td> 1,5</td><td></td><td>pSEC</td>

X

Table 12

<td>Symbol</td><td>Specification</td><td>Mine</td><td>Max</td><td>Unit</td>

<td>TD6</td><td>block transfer period</td><td></td><td> 750</td><td>pSEC</td>

<td>TD7</td><td>CCU data response time</td><td></td><td> 1,25</td><td>pSEC</td>

<td>TD8</td><td>VC response time</td><td> 1,25</td><td> 15</td><td>pSEC</td>

<td>TD9</td><td>handshake delay</td><td></td><td> 15</td><td>nsec</td>

<td>TDLo</td><td>VC block period delay</td><td></td><td> 150</td><td>pSEC</td>

<td>TH4</td><td>control data hold</td><td></td><td></td><td>nsec</td>

<td>Th5</td><td>status data hold</td><td></td><td></td><td>nsec</td>

<td>TH6</td><td>RC data hold</td><td></td><td></td><td>nsec</td>

<td>TS4</td><td>control data setup</td><td></td><td></td><td>nsec</td>

<td>TS5</td><td>status data setup</td><td></td><td></td><td>nsec</td>

<td>TS6</td><td>TC data setup</td><td></td><td></td><td>nsec</td>

<td>tw4</td><td>write width</td><td></td><td></td><td>nsec</td>

<td>tw4</td><td>read width</td><td></td><td></td><td>nsec</td>

<td>TW6</td><td>block request width</td><td> 1,5</td><td></td><td>pSEC</td>

FIG. 19A and 19B show, for PSK modulation at 16 levels, the timing relationships between the various transmit and receive speech blocks transmitted between VCU 17, 28 and CCU 18, 29. Top of FIG. 19A shows the system 25 star timing to which all transfers are assigned. The same raster timing applies to FIG. 19B. Each modem master has a length of 45 ms and includes four speech intervals (or channels). Each speech interval consists of two system voice block periods (SVBP) of speech30 data, each containing 82 symbols (requiring 5,125 ms) as well as an additional 16 initial data symbols requiring 1 ms of raster duration.

For the transmit channels, a 328 bit (41 bytes) block of processed speech is transferred from VCU 17, 28 to CCU

18, 29 prior to the beginning of each SVBP period during

/.

and voice codec block transfer period (VCBTP). It is seen that the 64 kbps input data stream of the VCU, which is associated with a processed speech block, is divided into VCBP periods of 22.5 ms. For example, the transmitting channel 0 in FIG. 19A, unprocessed VC input data in the VCBP periods OA1 and OBI are associated with processed data in the VCBTP periods 0A1 and OBI. It is also noted that the VCBP periods for channels 0 and 2 are set off half of a VCBP period (i.e. 11.25 ms) from the VCBP periods for channels 1 and 3.

For the receiving channels (cf. Fig. 19B), a 328 bit (41 bytes) block of processed speech is transferred from CCU 18, 29 to VCU 17, 28 at the end of each SVBP period in a VCBTP period. As in the broadcast channels, the time offset for the VCBP period relative to the VCBTP period depends on the implementation, and FIG. 19B shows a maximum displacement of one VCBP period. To understand the relationship between the input and output data of the speech codecs, reference is made to FIG. 19A and 19B. For the receiving channel 0, the compressed voice data transmitted during VCBTP OA10 and OBIO is linked to the processed, expanded data stream in VCBP OA10 and OBIO.

TCADDR lines 90 transmit broadcast channel address signals from CCU 18, 29 to VCU 17, 28. These three address lines are used to select the address of the broadcast channel concerned.

transmits transmit channel data signal CCU 18, 29.

transmits a signal for transmitted channel data from VCU 17, 28 to CCU 18,

29th The TCDAV signal indicates to CCU 18, 29 that a data byte is available in the TCDATA register. The TCDAV signal remains at a low level until the TCDACK signal is triggered.

The TCDACK line 93 transmits a transmit channel's data35 recognition signal from CCU 18, 29 to VCU 17, 28.

The TC data bus 91 crosses between VCU 17, 28

TCDAV line 92

/.

The TCDACK signal ports the data into the TC data bus and resets the TCDAV.

TCSCWR line 94 transmits a transmit channel status / control write signal from CCU 18, 29 to VCU 17,

28th The TCSCWR signal enters the speech codec control word into the appropriate transmit channel control register determined by the TCADDR lines. The data is loaded into the register on the leading edge of the TCSCWR signal.

TCSCRD line 95 transmits a transmit channel's station 10 000 / control read signal from CCU 18, 29 to VCU 17,

28th The TCSCRD signal port controls the status byte of the TCDAT Abuse from the voice codec status register designed through the TCADDR lines.

BLOCKRQ line 96 transmits a block request15 signal from CCU 18, 29 to VCU 17, 28. The BLOCKRQ signal is used to initiate the transmission of a 41 bytes data block from the voice codec (specified through the TCADDR lines) to CCU 18, 29 over the TCDATA bus . Talecodec uses the BLOCKRQ signal to start VCBP timing.

TCVCRST line 97 transmits a broadcast channel speech codec reset signal from CCU 18, 29 to VCU 17, 28.

The send voice codec specified by the TCADDR lines is reset.

RCADDR lines 98 transmit receiving channel address signals from CCU 18, 29 to VCU 17, 28. These address lines are used to select the address of the receiving channel as follows.

The RCDATA bus 98 transmits receiving channel data signals between CCU 18, 29 and VCU 17, 28.

RCDAV line 100 transmits a receive channel signal for available data from CCU 18, 29 to VCU

17, 28. The RCDAV signal indicates to it through

The RCADDR lines specified the speech codec that a data byte is available in the RCDATA register. The RCDAV signal port controls the data in the RCDATA bus and in the RCDATA register and resets the RCDACK line.

RCDACK line 101 transmits a receiving channel data recognition signal from VCU 17, 28 to CCU 18,

29th The RCDACK signal indicates to CCU 18, 29 that the data has been read from the RCDATA register and that another byte can be transmitted from CCU 18, 29.

RCSCWR line 102 transmits a receive channel status / control write signal from CCU 18, 29 to VCU

17, 28. The RCSCWR signal enters the control word into the appropriate speech codec10 control register designed through the RCADDR lines. The data is loaded into the register on the leading edge of the RCSCWR signal.

RCSCRD line 103 transmits a channel's status / control read signal from VCU 17, 28 to CCU 18, 29. The RCSCRD signal ports the voice codec status word into

The RCDATA bus from the status register designated by the RCADDR lines.

BLOCKRDY line 104 transmits a block ready signal from CCU 18, 29 to VCU 17, 28. The BLOCKRDY signal is used to initiate the transmission of a 41 bytes data block from CCU 18, 29 to the voice codec specified by the RCADDR lines. The voice codec uses the BLOCKRDY signal to initiate VCBP timing. CCU 18, 29 must have a data byte available in the preceding RCDATA register. the leading edge of the BLOCKRDY signal.

RCVCRST line 105 transmits a receiving channel's speech codec reset signal from CCU 18, 29 to VCU 17, 28. The speech codec specified by the RCADDR lines is reset by the RCVCRST signals.

The receiving channel's VCU hardware receives blocks of input data of 41 bytes from CCU 18, 29 during a VCBTP 'period as shown in FIG. 20A. After processing the data according to the operating mode in question, the 8 bit μ compressed data is transmitted at a frequency of 8 kHz to the PBX (STU) interface module. It is done in

VCU 17, 28 data buffer processing to simplify the input / output requirements of CCU 18, 29. Control information is transmitted between VCU 17, 28 and CCU 18, 29 via a set of control and status ports for each receiving channel at the beginning of a VCBTP. period, cf. 18. The following operating modes are performed by receiving codecs:

In external mode, bandwidth expansion is performed with an input data rate of 14.6 kbps (328 bits every 22.5 ms) to an output data rate of 64 kbps. The voice data may also include DTMF tones.

In internal mode, previously compressed is transferred

14.6 kbps speech from CCU 18, 29 to VCU 17, 28 through PBX 15 or STU 27. Since PBX 15 or STU 27 expects 64 kbps data, data streams must be completed. The 64 kbps output data consists of a pattern of empty bytes (FF hex) until the speech data becomes available from CCU 18, 29. Then a sight byte (55 hex) is output, followed by the previously processed 41 data bytes, followed by the empty byte pattern continues. FIG. 20A shows an example of input and output data timing and content for 16 PSK modulation.

In silence mode, input blocks of voice data are from

CCU 18, 29 consumed but not used. An outgoing blank byte pattern (FF hex) for PBX 15 or STU 27 is maintained to ensure line style.

In standby mode, continuous routines for hardware diagnostics are performed and the resulting status is stored in the status register. The transfer of blocks to CCU 18, 29 does not take place until operating mode is changed by a block request similar to VCBTPA. The new control word (and operating mode) is read by the voice codec and the diagnostic status information is transmitted to CCU 18, 29.

The transmit channel's VCU hardware receives 8 bit uncompressed PKM signals (sampling rate of 8 kHz) from

PBX / STU interface. After processing the data in henK 175353 B1 hold to the current operating mode, the data is delivered to CCU 18, 29 in blocks of 41 bytes during the VCBTP period, cf. 19A. Data buffer processing is performed in VCU 17, 28 to simplify the input / output requirements of CCU 18,

29th The control information is transmitted between VCU 17, 28 and CCU 18, 29 via a set of control and status ports for each transmit channel at the beginning of a VCBTP period, cf. 17. The following operating modes are controlled by send codecs:

In external mode, speech 10 bandwidth compression is performed with an output data rate of 14.6 kbps (328 bits every 22.5 ms). The processed speech data is transferred to CCU 18, 29 in the form of blocks of 41 bytes. The voice data may also include DTMF tones.

In internal mode, previously processed voice data is transferred from PBX 15 or STU 27 to CCU 18, 29 through VCU 17, 28. The arriving data stream of 64 kbps comprises a pattern of idle bytes (FF hex), a synchronization byte (55 hex), the 41 previous processed, compressed voice data bytes, and additional idle bytes until the next synchronization byte. The speech codec looks for the synchronization bytes in the arriving data, which takes place at the boundary of a byte, and takes over (buffer) the 41 bytes of speech data. Then the speech block is transferred to CCU 18, 29 during the next one

VCBTP period as described above. FIG. 20B shows an example of timing for input and output as well as data content for 16-PSK modulation. Section 1 of the output channel is a synchronization byte and Section 2 is a processed byte. The shaded sections represent an idle 30 byte pattern. It is noted that sight and speech data bytes do not occur across VCBP boundaries.

In silence mode, arriving voice data for PBX 15 or STU 27 is consumed but not used. The 41 bytes of outgoing voice data for CCU contain a quiet (silent) speech pattern.

In standby mode, continuous routines for hardware diagnostics are performed and the result is stored in the status register. Block transfers to CCU 18, 29 do not occur until operating mode is changed by a block 5 maturation similar to VCBTPA. The new control word and mode of operation are read by VCU 17, 28, and diagnostic status information is transmitted to CCU 18, 29.

The implementation of the RELP algorithm requires that a codec raster be defined, which must be an entire sub-multiple of the VCBP period, which is 22.5 ms.

As PBX 15 and STU 27 operate asynchronously with respect to the system's internal timing, VCU 17, 28 must include means for detecting, reporting and compensating for data overflows and overflows. This happens approx. once in 5000

VCBP-periods. While over / under detection is dependent on implementation, such errors are reported in the status word. In the case of data overflow, compensation can be made by repeating the last speech sample, while in the case of overflow one or more speech 20 samples can be ignored.

After resetting the codec circuit (s), VCBTPA will be the first block transmitted from CCU 18, 29, cf. 19A.

Channel Controller (CCU).

The channel controller performs similar functions in the subscriber stations and the base station. The hardware used for the CCU function in both types of station is actually the same. Software at the subscriber station deviates slightly from software at the base station. The CCU performs many functions related to formatting and timing of information related to the operation of the TDM channels. The CCU receives signals from four sources. These are first of all the digitized samples to be broadcast. These are transferred from VCU 17, 28 to / ·

CCU 18, 29, cf. 2 and 3. These data may be coded speech samples or data samples from the RS-232 data port 10 of the STU (Fig. 12). In any case, the digital channels operate at 16 kbps. Four channels can be handled simultaneously in the CCU unit 18 in the base station when all four transmission channels operate red 16 PSK modulation. The subscriber station CCU 29 operates on a single stream only, but this stream may be located at any of the four interval positions associated with the TDMA grid scheme. Next, it is input from the STU 27 in the subscriber station or the RPU 20 in the base station via the BCC channel to the CCU. These are control messages regarding operating mode, status and control information. Many of the BCC messages from CCU device 18, 29 are RCC messages received by CCU device 18, 29. CCU device 18, 29 transmits control information from RCC messages to STU 27 or RPU 20 and which replies receive control20 messages from RPU 20 or STU 27. This determines what CCU unit 18, 29 should do with the data from VCU 17, 28. The third input source provides timing and status information from modem 19, 30a. Modem 19 outputs the master clock signal used in

VCU-CCU-modem chain. Furthermore, the modem 19, 30a provides status on the accuracy of bit tracking, synchronization, RF AGC level adjustments and other proper functioning indications used by the CCU 18, 29 to determine if there are reliable communications across the channel.

CCU 18, 29 attempts to control the fine tuning of modem 19, 30a operation through transmit power change, AGC level and timing / distance measurement orders. Modem transmission quality metrics are reported to RPU 20 or STU 27. The fourth input source is the actual modem data received as symbols of up to 4 bits each (depending on modulation levels). These symbols are inserted into buffer circuits, demultiplexed, and the output signal is applied to the receiving circuits of the VCU 17, 28 for decoding.

FIG. Fig. 21 shows a block diagram of the CCU.

This CCU is designed as two one-way DMA data channels with intelligent microprocessor controller. The function of the DMA channels is to transfer the data from the VCU to the modem and vice versa. The CCU interface for the VCU comprises two parallel DMA buses, namely a TX bus 107 for the transmit channel (VCU-to-CCU-to-modem) and an RX bus 108 for the receive channel (modem-to-CCU-to-modem). VCU). The data processed by the transmit circuits in the VCU device is inserted into buffer circuits in the VCU storage until the CCU requests a

DMA transfer. In each block transfer period, 41 bytes are transferred to the CCU. Two of these blocks are transmitted per active voice channel (up to four voice channels in the base) per TDMA raster. The CCU receives these transmitted bytes through a transmit voice codec interface module (TVCIM) 109 and stores this data in a transmit memory module (TMM) 110. Depending on the specific mode of operation for the given channel, a CCU processor built into a microcontroller module (MCM) 111 provides the control / sync preamble to the encoded speech bytes, thus forming a complete speech packet for transmission to the modem through modem interface module 112. MCM module 111 maintains raster timing information and transmits the data to the modem at the correct time. Before transferring to the modem, the transmit data of the MCM module 111 is converted from the byte format used in the CCU to a symbol format of 1, 2 or 4 bits per second. symbol, depending on the modulation levels in that range.

The reverse process takes place for the receiving data35 one from the modem. The data from the modem is received in a modem 175353 B1 '96 deminterface module (RMIM) 114 and inserted into a receive memory module (RMM) 115. Then the data from the format used in the modem is converted at 1, 2 or 4 bits per second. symbol for the byte format used in the CCU unit and in all other baseband processing. MCM module 111 removes the initial bits and control bits from the data stream arriving over the RX bus 108 on the basis of raster timing knowledge, which the modem announces a raster timing module (FTM) 116, and on the basis of the modem's own identification of different passwords in the symbol stream. The converted data is fed to the VCU through a receive voice codec interface module (RVCIM) 117.

The CCU also creates link-level control of the RCC transmissions at both the base station and the subscriber stations. In the base station, there is only one CCU device - the RPU device - for processing the RCC channel. The CCU controls the reception and formatting of the messages from the base station RPU to the STU control circuit in the subscriber stations. The CCU control unit 20 includes detection and error checking in the RCC messages as well as formatting and packet design of the RCC information for transmission over the radio connection. The CCU also detects collisions in the arriving RCC message at the base station. The CCU controls the power and distance calculations of the subscriber stations that are undergoing capture attempts. The capture protocol and other RCC functions have been described above.

FIG. 22 shows the functional, software-implemented design of the CCU unit. The CCU includes three separate data paths: TX TX 107, Receiver Bus RX 108 and Microcontroller Local Bus 119. Microcontroller 111 shares TX Bus 107 with a DMA Controller 120 and shares RX Bus 108 with a DMA Controller. 121st The microcontroller 111 uses these buses to control the peripheral circuits of the DMA controller, the control / status registers 122 and to access the send buffer memory 110 and the receive buffer memory 115. The control / status registers 122, branched from the microcontroller's local bus 119, create an interface to the RFU device, modem and CCU hardware. An RS-232C link 123 between RPU and CCU is supported by a UART on the microcontroller chip 111. In the subscriber station, the RPU is replaced by the STU, but the interface remains the same.

The microcontroller 111 has access to three physically separate RAM areas: local RAM, send buffer, and receive buffer. The local RAM can be divided into on-chip RAM and off-chip RAM. The microcontroller can only access the send buffer and receive buffer when the DMA controller in question is idle.

The transmission buffer 110 is divided into a number of separate segments. Each segment contains the skeleton of a voice or RCC packet ready for transmission over the channel. The preamble and the unique word (RCC only) are constants that the microcontroller 111 initializes after the CCU reset. The password (speech only), voice data, and RCC data are input by the microcontroller into transmit buffer 110 immediately prior to the DMA transfer to the modem

19, 30a. Since the RCC null ACK is a fixed message transmitted at high frequency, it is stored in a separate unit in the transmission buffer 110.

Receive buffer 115 is divided into a number of separate segments. One segment serves to store the voice data, which is inserted into buffer and transmitted in VCU block form. The RCC data is inserted in buffer separately from the speech data so that it is stored for a longer period of time. If necessary, the microcontroller 111 in the receive buffer 115 can maintain two rasters of the RCC sequence, thus making the RCC copying task (from buffer to local RAM) less time-critical.

The local RAM contains the work variables used by the microcontroller 111. An important data structure stored therein supports the BCC channel between CCU and RPU. A register in the local RAM has the function of creating queue information for the RS-232C interrupt handler. An arrow and a field in this register part define the active transmit data block (TXDB) from which data is read and transmitted. TXDB contains length and arrow information in the next TXDB in the queue, creating a linked list. On the receiving side, a circular buffer is used to store arriving data bytes. When a complete message is received, the interrupt handler sets a flag in the serial code for interpretation.

Microcontroller 111 uses the local bus

119 to access the modem and to the RFU and CCU control / status registers 122. The bus also creates access to the TX bus 107. and RX bus 108 through intermediate logic circuits 124 and 125, respectively. For the avoidance of contention, microcontroller 111 only accesses buses 107 and 108 when the DMA controller 120 or 121 in question is idle.

CCU and RPU communicate via connection 123 through a full-duplex RS-232C interface, namely the BCC unit. The asynchronous characters of binary 8 bit code are broadcast on

9600 Baud. One frame bit and one stop bit are used to frame data bytes. The messages end with a unique byte, and fill bytes are used to prevent the unique byte from occurring within a message. To ensure connection integrity, an alternate30 bit protocol and an 8 bit checksum are used.

The microcontroller provides two external interrupts. One is provided by the DMA transmitter controller 120 and the other by the DMA receiver controller 121. These switches occur when the controller 120, 121 terminates its biocoverageDK 175353 B1 guidance. This leaves the control of the particular bus to the microcontroller 111.

The BCC interface is controlled by an internal switch. Software is interrupted upon receiving or transmitting a byte.

At the base station, CCU microcontroller 111 is responsible for controlling and monitoring the entire assigned four-channel data path, including VCU 17, 28, CCU 18, 29, modem 19, 30a, and RFU 20, 31a. In the subscriber station, the microcontroller 111 manages and monitors the same hardware, but operates only one data path. The CCU is controlled by the RPU (in the base station) or the STU (in the subscriber station).

The CCU provides the VCU with operating mode information. The mode changes occur only at the limits of system range. During the speech compression operation, the CCU also provides the VCU with information about the position of the VCU block in the system interval (there are two VCU blocks per system interval). The VCU20 addressing is established by the CCU device prior to a data transfer, thereby performing the MUX / DEMUX task. VCU status is read by the CCU after each block transfer and appropriate statistics are maintained in the CCU. The CCU can also pre-reset VCU hard ware.

The microcontroller 111 delivers the current modulation level to a symbol-to-byte converter 126 via RX bus 108 and to a byte-to-symbol converter 127 via TX bus 107.

The modem is provided with information regarding the type of data received, RCC or speech, depending on the different capture procedures used for their reception. The modem notifies the CCU of each interval of a fractional clock offset, AGC35 level, and connection quality value. CCU frequencyDK 175353 B1

100 the assignment is specified by the RPU or STU device. The CCU provides modem reset (so-called modem hard reset), self-test or training mode on the receiving side.

The CCU processes the data stream full-duplex through the transmit and receive buses 107 and 108. During a given interval time, the transmit data from the VCU is transmitted blockwise to the transmit buffer 110 via the DMA transmit controller 121. Each block has a length of one VCU block, so two such transfers are required for each voice channel. The CCU provides the VCU with the appropriate channel address prior to transmission, thereby performing multiplexing.

In the transmission buffer 110, a preamble and password transmitted prior to the VCU data are stored at the beginning of each interval. The DMA unit transfers the transmit data from the transmit buffer to the rate-controlled FIFO stack storage 128, while the modem receives data from the FXFO stack storage 128 as needed. The byte-to-symbol conversion is performed in converter 127 during transfer. The microcontroller provides control of the peripheral circuits of the DMA transmitter together with generation and insertion of the speech packet password.

In fact, the data stream on the receiving page is almost a mirror image of the data stream on the sending page. The data from modem 19, 30a is loaded into the rate-controlled FIFO stack storage 129. DMA unit controls 121 clear the FIFO stack storage 129 in the receive buffer 115 as needed. Symbol-to-byte conversion takes place in converter 126 and raster timing is performed by clock circuit 130.

The coincidence boundaries are obtained automatically when the channel is synchronized. When a complete VCU block has been received, the DMA unit transfers the block to that VCU. The microcontroller 111 provides control of the receiving DMA controller.

Password detection is performed for each interval. The microcontroller 111 performs this task by / ·

101 copy the password byte into the local RAM and by comparing this word with a list of valid passwords. In each interval, the modem 19, 30a provides a time delay (fractional symbol offset) and an AGC value. Microcontroller 111 reads these sizes and makes an appropriate interpretation. If there are problems with power and distance, the subscriber station is notified via the transmitted password.

The CCU performs synthesizing the transmitted RCC data in the sending buffer 110 according to the content of the RCC message queue. If the RPU has sent an RCC message to the CCU, the message is formatted in send buffer 110. Otherwise, the NULL KNOWLEDGE message stored in send buffer 110 is used. When the RCC package is ready, the RCC preamble, the unique word and the RCC data are transferred to the modem 19, 30a as needed. The CCU performs collision detection and sets up the corresponding, outgoing detection bit for RCC collision.

The circuit that processes the received RCC data operates in two modes: frame search and monitor. In grid search mode, the RCC channel is considered to be out of sync. Each arriving RCC message must be synchronized using a unique word detection algorithm. In monitor mode, the RCC channel is synchronized and this algorithm for searching for the unique word is not used. The base station is always in grid search mode, since subscribers can come in at the wrong time at any time. In the subscriber station, the RCC data processing circuit is always in monitor mode, unless the station has not achieved RCC synchronization.

In grid search mode, the unique word (UW) is detected after each RCC interval. Microcontroller 111 performs this task by performing scanDK 175353 B1

102 call for the unique word in a window around the nominal location of the unique word. When detecting the unique word, the CCU is provided with symbol timing information.

The received RCC data is DMA transmitted from the modem

19, 30a to the receive buffer 115. When the transfer is complete, the RCC data is copied into the local RAM storage for processing. The received RCC packets are filtered by the CCU device. An RCC package is only transferred to

The RPU if the unique word is detected and the CRC is correct.

During the RCC operation, the corresponding VCU channel is in standby. During this period, no data is transmitted between VCU and CCU, neither over the transmission data path 107 nor the receive data path 108.

Software is performed with an Intel 803 microcontroller 111. The program storage is located in an external EPROM storage on the microcontroller's local bus. Software is required to respond to requests for DMA service on a timely basis while maintaining a 64 kbps data stream in both directions without loss of data. A FIPO buffering in the stack bearings 128 and 129 on the modem interface provides the microcontroller 111 with the necessary time to perform the DMA block transfers and system control functions.

The software is divided into five separate modules: supervisor, data transfer, BCC transceiver, BMM control and utility. Each module is designed to have only a single entry point and a single starting point, except for switches and failure modes. A further exception is the utility module, which contains various utility routines to which the other modules have direct access. In general, communication between modules takes place through the use of global variables that are defined in a sepa- rate data segment.

103

The Supervisor module includes an initialization function that maintains global program control and performs basic self-test functions.

The data transfer module provides data transfer control over the TX bus 107 and RX bus 108 for both the voice data and the RCC data, performs the synchronization word detection at all modulation levels for both the voice and RCC data, and manages the CCU-RPU RS-232 communication connection 123rd

The BCC transceiver module performs BCC transceiver tasks, processes the BCC queues, formats sent BCC messages, processes received BCC data, and moves the RCC data in and out of the CCU through the BCC channel.

The BBM control module manages RFU, modem, VCU and CCU hardware via registers, reads and interprets status information from these constituents (eg modem AGC, connection quality and symbolic ambiguity), decodes entered passwords in the receive voice channel, formats the passwords channel channel passwords, real-time software / hardware timing and performs self-tests online.

The utility module performs various utility routines that other modules have access to.

The CCU software is divided into four separate processes that operate in the main proceedings in a continuous manner. Three of them are BCC data, TX DMA and RX DMA processes that are interrupted and only used when a specific event requires it. All three event-driven processes are in the data transfer module. The final process, distributed among all modules, is a background process that initializes, manages and monitors the other three processes.

BCC messages from the RPU (or the STU at the subscriber station) are received and buffered by the BCC data process. When a complete announcement has been made

/.··

104 received, the BCC data process notifies the background process via a mailbox. The background process examines the mailbox under the main loop so that any new message is detected. The background process interprets the messages and takes appropriate precautions. Any responses are loaded by the background process in the queue of BCC broadcast messages, and the BCC data process is notified.

BCC messages may require a reshaping of

CCU data channels. The required control information is loaded into the modem 19, 30a and the VCU 17, 28 at appropriate times. The modem responds to a new control word at the interval limits. The VCU expects mode changes to occur on the first VCU block transfer at an interval limit. The background process is responsible for maintaining correct control timing.

The collection of status information is done by the background process, the TX DMA process and the RX DMA process. The latter two collect status words from the TX side and the RX20 side of the VCU. This is necessary because these status registers are accessed only via the TX bus 107 and the RX bus 108, which are only idle for limited periods of time. The background process collects status information directly from modem 19, 30a through the status registers 122 of the local bus 119. The background process collates all the collected status information stored in specific status variables. The background process processes status requests from the RPU based on this collected status information.

Certain status information such as AGC value and time offset may require action from the CCU. In addition to being stored as status information, such data is used to correct subscriber station power and distance issues. In the case of RCC messages, information on power and AFK 175353 B1 is entered

105 stand directly to the RPU as part of the RCC channel. The background process performs this function by formatting a BCC message containing RCC data. AGC and distance. When the packet is ready, it is placed in the BCC sending queue and the BCC data process is notified. For voice channels, this status information is used to format passwords that are embedded in outgoing voice packets. The background process performs this formatting function and controls the transmission of the password through the voice channel. All passwords must be sent five rasters in a row, thereby obtaining a coding with redundancy 5/1. The TX DMA process automatically sends out the password selected by the background process.

The background process also maintains real-time software / hardware clock management. This is done by querying one of the 8031 ​​unit's hours and by counting overflows. The real time clock management creates a time base for software timeouts and other time-dependent events. The background process checks that system timing is maintained20 by collecting information from the CCU hardware error indicators and by checking that the data transfers are taking place correctly in the system grid. The system raster information is obtained over the start of the system raster status line and with a timer connected to the 16 kHz clock generator 130. The data synchronization is performed by the background process.

The BCC data process responds to RS-232 interrupts, which can occur both in the transmitter direction and in the receiver direction in the port. The system is just to output another byte on the sending page or receive another byte on the input page. An end-of-message indication on the receiving page causes the BCC data routine to notify the background process.

The TX DMA process and the RX DMA process respectively control the DMA transmit channel and the DMA receive channel.

/.·.

106

The steps below are indicated below. step description of the data transfer function controlled by the software. The events in the data transfer process are marked by DMA controller interruptions. The interrupt occurs after the DMA5 controller completes the assigned block transfer. Each review starts with the beginning of an interval data transfer. For that part of the specification, it is appropriate to look at FIG. 23 and

24th FIG. 23 is a timing diagram for transmitting RCCs and

16 PSK voice data over the CCU unit's sending bus. FIG. 24 is a timing diagram for transmitting RCC and 16 PSK data over the CCU unit's receiving bus. Table 13 and Table 14 indicate the characteristics of the 23 and 24, respectively, shown time symbols.

Table 13

Time Typically

<td colspan="2">Symbol</td><td>Operation</td><td>Max (ps)</td><td>Min {ps)</td><td>(WS)</td>

<td></td><td>CCU</td><td>DMA setup</td><td> 150</td><td> —</td><td> 100</td>

<td>^HVCB</td><td>VCU</td><td>DMA transfer</td><td> 600</td><td> -</td><td> 100*</td>

<td>^fcRCC</td><td>RCC</td><td>overførinq</td><td></td><td></td><td></td>

<td></td><td>from</td><td>CCU</td><td> -</td><td> -</td><td> 900</td>

<td>^fcM0</td><td>RCC</td><td>tx modem block</td><td> -</td><td> 10350</td><td> 10350</td>

<td>^HM2</td><td>1st</td><td>Rx modem block</td><td> -</td><td> 4300</td><td> 4300*</td>

<td>^HM3</td><td>2nd</td><td>Rx modem block</td><td> -</td><td> 4225</td><td> 4825*</td>

* Based on RELP VCU.

107

Table 14

Time Typically

<td> 5</td><td>Symbol</td><td>Operation</td><td>Max (ys)</td><td>My (vs)</td><td>(US)</td>

<td></td><td>^HS</td><td>CCU DMA setup</td><td> 150</td><td> -</td><td> 100</td>

<td></td><td>^tVCB</td><td>VCU DMA transfer</td><td> 600</td><td> . -</td><td> 100*</td>

<td></td><td>SLO</td><td>1st Tx modem block</td><td> -</td><td> 5225</td><td> 5825*</td>

<td> 10</td><td>Sil</td><td>2nd Tx modem block</td><td> -</td><td> 4225</td><td> 4825*</td>

<td></td><td>* M2</td><td>RCC Rx modem block</td><td> -</td><td> 5600</td><td> 5800*</td>

<td></td><td>^tRCC</td><td>RCC transmission</td><td></td><td></td><td></td>

<td></td><td></td><td>to CCU</td><td> -</td><td> -</td><td> 900</td>

* Based on RELP VCU.

Transmission Function - RCC

First Receive end of TX DMA transfer interrupt. This signals that the processing of the previous interval has ended and that the processing of the next interval can be started. The TX DMA process is called.

a. Print channel and modulation switch information. This information is required by modem 19, 30 and byte-to-symbol converter 127.

b. Format any pending RPU RCC message in send buffer 110. Otherwise, prepare and send the NULL-ACK message.

c. Initialize and unlock DMA transfer from send buffer 110 to modem 19, 30a with arrow on RCC preamble, unique word and RCC data block.

d. Return from interrupt and continue with background processing.

·' 108

Transmit function - speech

First Receive end of TX DMA transfer interrupt. This signals that the processing of the previous interval has ended and that the processing of the next interval can be started. The TX DMA process is called.

a. Print channel and modulation switch information for the next interval. This information is required by modem 19, 30a and byte-to-symbol Vr&gt; n ”erder 127.

b. Select VCU port address and unlock DMA transfer from VCU to send buffer 110.

c. Print VCU control words.

d. Disconnect VCU to start transferring.

e. Return from the interrupt and proceed with the background process.

2nd Receive end of TX DMA transfer interrupt. This signals that the transfer from the VCU to the send buffer is completed. The TX DMA process is called.

a. Read VCU status words.

b. Enter the password for the sending buffer 110.

c. Initialize and unlock DMA transfer from send buffer 110 to modem 19, 30a with arrow on the speech preamble, password, and speech data block.

d. Return from the interrupt and proceed with the background process.

Third Receive end of TX DMA transfer interrupt. This signals that the transfer of the first half of the interval from the sending buffer 110 to the modem 19, 30a is completed. The TX DMA process is called.

a. Select VCU port address and unlock DMA transfer from VCU to the send buffer.

109

b. Print VCU control words.

c. Disconnect VCU to start transmission.

d. Return from the interrupt and proceed with the background process.

4th Receive end of TX DMA transfer interrupt. This signals that the transfer from VCU to send buffer is complete. The TX DMA process is called.

a. Read VCU status words.

b. Initialize and unlock DMA controller 120 for the routing transfer buffer.

c. Return from the interrupt and proceed with the background process.

Receive Function - RCC

First Receive end of RX DMA trans fer interrupt. This signals that the processing of the previous interval has ended and that the processing of the next interval can be started. The RX DMA process is called.

a. Prepare for BPSK modulation. This information is required by the symbol-byte converter 126. At this point, the modem 19, 30a has already received this information.

b. Initialize and resolve DMAr Transfer from mo25 demet 19, 30a to receive buffer 115 for the RCC message.

c. Return from interrupt and proceed with background processing. The AGC calculation and the processing of ambiguity in bit synchronization should take place at this time.

2nd Receive end of RX DMA transfer interrupt. This signals that the RCC transfer from modem 19, 30a to receiving buffer 115 is complete. The RX DMA process is called.

a. Copy the RCC into the local RAM storage.

110

b. Return from the interrupt and proceed with the background process. Prepare transferring received RCC to RPU if the unique word is detected and the checksum is correct.

Receive function - speech

First Receive than by RX DMA transfer interrupt. This signals that the processing of the previous interval has ended and that the processing of the next interval can be started. The RX DMA process is called.

a. Provide speech data with proper modulation. This information is required by the symbol-offer converter 126. At this point, the modem has already received this information.

b. Initialize and unlock DMA transfer from modem 19, 30a to the receive buffer for the first half interval of the voice data.

c. Return from the interrupt and proceed with the background process. The AGC calculation and processing of ambiguity in bit synchronization and of password should take place at this time.

2nd Receive end of RX DMA transfer interrupt. This signals that the transfer of the first half of the interval from modem 19, 30a to the receive buffer 115 is completed. The RX DMA process is called.

a. Select VCU port address and unlock DMA transfer from receiving buffer 115 to the VCU. Disconnect VCU to start transferring.

b. Return from the interrupt and proceed with the background process.

Third Receive end of RX DMA trans fer interrupt. This

111 signals that the transfer of the first half of the interval from the receiving buffer 115 to the VCU is completed. The RX DMA process is called.

a. Initialize and unlock DMA controller 121 to transfer the second half of the range from modem to receive buffer.

b. Return from the interrupt and proceed with the background process.

4th Receive end and RX DMA transfer interrupt. This signals that the transfer of the second half of the interval from modem 19, 30a to receiving buffer 115 is completed. The RX DMA process is called.

a. Select VCU port address and unlock DMA transfer from receive buffer 115 to VCU. Disconnect

VCU to start transferring.

b. Return from the interrupt and proceed with the background process.

The CCU software execution.

The execution of the program begins as the result of hardware reset and the sequence of operations begins in the supervisor module. This module takes care of any hardware and software initialization before entering a main loop. This supervisor module performs some basic self-test functions after hardware reset and at the request of the RPU. The main loop gets sequential access to the other modules. The supervisor module design is such that the tasks are divided into manageable time sections, which guarantees that the main loop exhibits a reasonable periodicity in the worst cases (so-called reasonable worst case periodicity). Tasks that require real-time response are handled by interrupt routines.

Each interrupt routine performs the minimum of beDK 175353 B1

112 action required to fulfill the service request. This is done as far as possible to maintain the serial nature of the program execution and to minimize interrupt queue formation. An interrupt routine will typically transfer data to and from an interface and perform a Boolean operation to indicate that the action has been performed. Then, one of the main loop addressed, serially executed code will process this information as required.

CCU microcontroller 111 is a data stream machine in so far as the software events are driven by the arrival and departure of data. Precise system timing creates the framework for this data stream, but the software events are derived directly from the data stream and not from the sy15 voice tags. This solution allows the software to respond to real events such as data in / Requests rather than artificial events such as system timing marks. The software relies on the hardware to transform the software's asynchronous actions into events that are synchronous with system grid timing. In order to do this, there must be certainty that things are initialized and ready before the system grid events occur.

It is therefore clear that the CCU software, when not heavily loaded, is expected to respond to incidents and perform certain tasks within a limited time. This real-time processing is driven by interrupts and therefore requires very careful design. There are four real-time incidents for the microcontroller that involve potential conflicts of danger, namely DMA Transmit Function, DMA Receive Function, RS-232 Transmit Function and RS-232 Receive Function. RS-232 interruptions have the lowest priority, occurring a maximum of once per second. ms. The software is designed to comply with this 1 ms limitation.

The response time for voice and RCC data handling is more critical, which is discussed in more detail below.

/...

113

The relative timing of the data transfers over the sending bus and the receiving bus is shown in FIG. 23 and 24. These diagrams are drawn approximately in scale and show timing in the worst case. The time multiplex nature of the sending and receiving buses is clearly shown in the diagrams. The dark cross lines shown on the sending and receiving paths correspond to microcontroller activity on the respective buses (t_g, t_RCC). During this time, the DMA controller concerned is 120, 121 idle. The short periods of time between the activities of DMA controllers (tycB ^ correspond to VCU block transfers. During this time, the DMA controller is engaged in that VCU. In the remaining time (t ^, t_M1, t_M2, t_M3) has the task of operating the DMA controller 120, 121 to operate the modem interface.

The clock-controlled FIFO stack layers 128, 129 in the modem interface create the primary, in the time diagrams, implicit time constraint. The FIFO stack store contains 16 symbols and creates a buffer time of 1 ms before discharge (TX) or overfill (RX). During this ms, the CCU unit can use the send or receive buses 107, 108 to perform the block transfers to and from the VCU or copy RCC data into the local RAM storage.

During switch-on, the CCU software performs an internal self-test and places the VCU, modem and RFU in their default state. The microcontroller 111 monitors the system grid timing and begins to perform block transfers to allow the VCU to synchronize. After initiating data transfers, the microcontroller 111 uses the DMA end of a block interrupt to maintain system timing. This interrupt is directly connected to the CCU unit's data throughput and thus to the 16 kHz symbol clock generator 130. The VCU unit maintains system timing implicitly through DMA transfer requests generated by the microcontroller 111 as a result of end-of-block interrupt. The microcontroller 111 continuesDK 175353 B1

114 monitor raster timing to ensure proper system operation is maintained.

On the subscriber station side, the system includes startup radio synchronization, which is performed by locating the radio control channel from which system timing is derived. Once the receive timing has been established, it is the job of the microcontroller Ill to determine the transmission timing to the base station.

The data transfer module takes care of the timing control and the background data transfer in the CCU. There are various data transfer functions for the sending data path, the receiving data path, the sending BCC device and the receiving BCC device. All of these tasks are controlled by interrupts that require real-time response. The module also performs sync capture and monitoring as a background task.

Transmission data transactions are called when the TX-DMA controller 120 requires service. This is typically done after a DMA bioconvergence, at which time for peripheral DMA circuits, end-of-block transfer interrupt is called, which interrupt is received over one of the two external interrupt lines on the 8031 ​​microcontroller 111. The service this interrupt requires depends on the type of data transfer, namely RCC or voice, and the time it occurs within the interval.

Sending data interrupt occurs at predictable times in each interval. FIG. 23 and 24 show the timing and duration of these interrupts. Each time, the microcontroller 111 must initialize the peripheral DMA circuits for the next block transfer. This operation should be performed within 150 ys from interrupt request to interrupt execution. In the case of RCC data, the first service request requires the microcontroller 111 to format the RCC message in

/...

115 the sending buffer 110 prior to DMA transmission. The operation must be completed within 900 ps. Since the operations over the transmitter are usually short and require prompt response, interrupt is given the highest priority.

The only output from send data path interrupter is VCU status words, collected after the VCU block transfer. The status word is analyzed with software in the BBM control module.

Receiving data paths are called when the receiving DMA controller 121 requires service. This typically occurs after a DMA block transfer, at which time the peripheral DMA circuits call an end-of-block transfer interrupt. This interrupt is received over one of the two external interrupt lines of the microcontroller 111 of type 8031. The service this interrupt requires depends on the type of data transmission, RCC or speech, and the time at which it occurs within the range.

Receipt data interrupt occurs at predictable times within each interval. FIG. 23 and 24 show the times and duration of these interrupts. Each time they occur, the microcontroller 111 must initialize the DMA controller 121 for the next block transfer. This operation should occur within 150 ps from interrupt request to interrupt execution, if DMA initialization is the only task to be performed. In the case of RCC data, the last service request is that after DMA transmission, the microcontroller 111 must copy the RCC message from the receive buffer 115 to the local RAM storage. This operation should also be performed within 900 ps. Because send route operation can occur within this time interval, receive route interrupt has lower priority than send route interrupt. Receiving data path interrupts make the VCU status word available after each VCU block transfer. This status word is analyzed with software in the BBM control module.

116

The handler also reads new RCC messages from the channel, which are interpreted in the BCC transceiver module.

The BCC receiving module is implemented via on-chip RS-232 UART, which is capable of providing one in5 interrupt that is triggered each time a receive or broadcast is received. BCC trades call a status bit to determine which of the two cases has caused interrupt, and on that basis it makes sure to activate that port.

The Baud clock generator is programmed to operate on 9600 Baud, with a maximum of 1920 interrupts per second. second. To avoid data loss, each interrupt must be processed within 1 ms. Since the typical interrupt frequency is low and the response time relatively long, BCC data transfer interrupt has low priority.

The BCC data transfer handler uses arrows to queue the received and transmitted data in line and out of line respectively. Here, only single-level processing is done, including byte filling and end-of-message insertion. These actions are described in the system interface specification.

There is very little data processing in the BCC transceiver module, whose main task is to put the data in and out of queue while handling the send, mod25 take and BCC data paths. The data synchronization capture and monitoring described below includes the main processing capabilities of the BCC transceiver module.

Sync word detection involves a symbol-level synchronization operation. The term syncord generally covers both the unique word in the RCC channel and the password in the speech channels. The Unique Word (UW) is a fixed 8-bit pattern placed at the beginning of an RCC message. A password (CW) is usually any of 8 possible 8-bit patterns located at the beginning of a voice channel. In addition to their synchronization

117 task, the passwords are used to indicate connection status, power adjustments and distance adjustments.

The base station's CCU unit must perform an exhaustive search for a valid RCC message at each interval. It performs this task by searching on the basis of system master timing for the unique word in a window of ± 3 symbols around the nominal UW location. The search algorithm begins with the nominal UW space and changes a symbol to the right and left until it

1) finds the UW pattern and 2) checks that the RCC checksum is correct. The search is completed as soon as 1) and 2) are met or when all options have been explored. The information about this shift, the RCC message, and the effect information is sent to the RPU after a successful search.

In each speech interval, the base station's CCU unit searches for a valid password in the received speech data. Only the nominal position of the password is checked, since no active symbol synchronization is performed during speech operation. If no password is detected in five successive rasters, the channel is declared out of sync and the RPU is notified of this condition. At this point, it is up to the RPU to take any appropriate action. The synchronism is said to have been restored when three out of five successive rasters successfully detected the password.

When receiving RCC data, the subscriber station CCU unit may be in one or the other of the following two modes: raster search or monitor (monitor). Grid search mode is used to capture receiver grid timing in the arriving RCC data, and it is automatically called when RCC synchronization is lost. Switching to monitoring mode when raster synchronization is achieved.

118

When in grid search mode, the subscriber station CCU must perform a comprehensive search for a valid RCC message after each RCC interval. Like the base station CCU unit, it performs this task by searching for the unique word in a window of ± 3 symbols around the nominal UW location, based on timing derived from the modem's detection of AM hole. The search algorithm begins with the nominal UW location and switches a symbol to the right and left until it 1) finds the UW pattern and 2) checks that the RCC checksum is correct. The search is completed as soon as 1) and 2) are met or when all options have been explored. The switching information after a successful search is used to adjust the receiver raster markings produced by the CCU. Capture ends when 1) and 2) are met for three successive rasters with the UW word in its nominal position. The STU will be notified of raster capture when this occurs. In grid search mode, the RCC messages are not forwarded to the STU.

When the raster capture is completed, the subscriber station CCU unit switches to monitor mode. Check only for the nominal UW location to avoid the possibility of false UW captures. If no UW word is detected in five successive rasters, the channel is declared out of sync and switches to raster search mode. The STU is notified of this unsynchronized state. In monitor mode, the RCC messages that have the correct checksum as well as the SIN30 number are transmitted to the STU.

In each voice interval, the CCU unit of the subscriber station examines the received voice data to find the correct password. Only the nominal password location is checked, since active symbol synchronization is not performed during the speech operation. In this rightDK 175353 B1

119 for the channel, all possible passwords are examined. The passwords can cause incremental changes in the power and distance values ​​of the subscriber station. Incremental distance changes can actually lead to symbol5 change as well as change in the values ​​of the distance determination. If no password is detected in five successive rasters, the channel is declared out of sync and the STU is notified. The synchronism is considered to be restored when, after three out of five successive rasters, successful detection of the password has been achieved.

Additional considerations regarding the CCU unit.

The request for transmit DMA transfer between transmitter buffer 110 and modem 19, 30a must be deduced from the total bit amount of FIFO stack storage 128, which means that FIFO stack storage 128 must always be filled when a DMA block transfer is complete.

The request for receive DMA transfer between mo20 demet 19, 30a and receive buffer 115 must be deduced from empty bits in stack storage 129, which means that FIFO stack storage 129 is always empty when a DMA block transfer is complete.

The CCU controller software creates gateways for DMA25 transfers, but there must be external controls to create handshaking to initiate and maintain block transfer. This is especially important for the modem interface, where raster timing is critical.

The microcontroller 111 must be capable of allowing a DMA transmission to wait. The software does not attempt to use the DMA bus during a block transfer unless this control is performed or the peripheral DMA circuits are idle.

The clock controlled FIFO stack bearings 128, 129 should be periodically and automatically reset.

120

The raster timing information must be available to the microcontroller 111. This could take the form of a symbol clock input for internal timer in the microcontroller.

When the CCU receives a synchronized RCCeller voice pack, no symbol change is required to bring the packet to the byte boundary. This should apply regardless of the modulation level.

modem

The modem operates in one of three operating modes. In the base station, the modem performs full-duplex sending and receiving function. At the subscriber station, the modem operates half-duplex, emitting under one part of the TDMA raster and receiving in another part of the TDMA raster. The third mode is a self-adapting training mode. A single modem design performs all three functions. The modem performs the appropriate function in response to key signals arriving from the controlling CCU device.

The modem 30a in the subscriber station and the modem 19 in the base station are identical, and FIG. 25 shows a block diagram of the modem.

The modem's transmit portions include a TX symbol filter 132, a digital-analog converter 133, a 200 kHz band-pass filter 134, a mixing step 135, and a TX timing control circuit 136. The modem's receiving portion comprises a mixing step 138, an analog-to-digital converter 139, a FIFO stack memory 140 and a TMS 320 microprocessor 141.

The modem transmitter transmits the information arriving from the CCU with 16-level PSK modulation. The CCU device on the receiving side should interpret the data as being on DPSK, QPSK or 16 PSK. The modem emits without knowledge of the modulation level.

The modem transmitter part is fully implemented in hardware and requires no adjustment. The symbols that

121 is received from the CCU, encoded, and the corresponding wave signals are formed to ensure appropriate interference characteristics and to avoid amplitude distortion and distortion by group delay. This ratio is enriched on the assumption that in the frequency band (within 50-100 kHz) closest to the band used there are no strong interfering signals (power densities of 30-40 dB below the signal). The transmitter portion of the modem uses a relatively wide intermediate frequency filtering (100 kHz) so that the transmitted signal is not subjected to amplitude distortion and group delay, and it also filters out the harmonics produced by the baseband digital filtering.

The TX symbol filter 132 is a digital FIR filter with fixed coefficients. This filter 132 simulates a six-pole filter with a sampling frequency of 50 samples per second. symbol per 6 symbols in the FIR filter.

The modem receives the symbols from the CCU unit in question at a rate of 16 kbps. These symbols are then converted into a DPSK code which is fed over the lead 143 to the FIR filter 132. The FIR algorithm requires that every other symbol be inverted before it is applied to the FIR filter. For DPSK encoding, a Gray code is used. This is intended to ensure that if an incorrectly attached symbol is received, there is a good likelihood that the two receive codec symbols have only one bit error.

The FIR filter 132 has a pulse response cut to 6T (T = 1/16 kHz). The FIR filter over-samples the symbols at a frequency of 800 kHz, so that each symbol is sampled 50 times during its 5T stay in the filter. This is equivalent to a sampling frequency of 3T / 25, where the sampling period is T / 25, so that the samples are delivered every 3T / 25 period. The output signals are offset so that there is only overlap between the first and the fourth, the second and the fifth at all times

/.

122 or the third and sixth pairs of samples. Each of these samples with a length of T / 25 is actually divided into two parts. In the first half of the sample period, the Id of the output signal is calculated, while the Q portion of the output signal is calculated in the second half of the period. Hereby, the current frequency at which the FIR filter 132 emits the data is 50 x 16 kHz »800 kHz. I and Q sampling are offset by half a sample period, but this is corrected by the FIR filter 132.

The signals representing the multiplication of the symbols and impulse responses in the FIR filter 132 and the addition of these two multiplications are created by an 8K.8 ROM memory which, depending on the symbols arriving over the line 143, is output over the line 144.

The FIR filter 132 outputs over the line 144 digital samples of 10 bits at a frequency of 800 kHz. These signals are applied to the D / A converter 133 for providing an analog wave signal over line 145. This wave signal is time-shared I and Q signals20 of the symbol to be emitted. This time-shared wave signal over line 145 is filtered into the 200 kHz bandpass filter 134 and then applied to the mixing step 135 via line 146. The mixing step is fed with a 20 MHz medium frequency local signal over line 147. This converts the I and O components to a 20.2 MHz medium frequency output over line 148. This output signal over line 148 is applied to a 20.2 MHz bandpass filter not shown. to the RFU unit 21, 31a.

The desired signal from the D / A converter 134 is centered at 200 kHz and has a bandwidth of approx. 32 kHz. By multiplying the 200 kHz wave signal by 20 MHz, the Iog Q samples are mixed with the SIN and COS components of the intermediate frequency. This allows the 20 MHz signal to directly output the 35-wave signal, and the exact component multiplication /.

123 tions are handled automatically. Therefore, there is no need for discrete generator circuits for SIN (IF) / COS (IF) to multiply the I / Q samples from the D / A converter as in the receiver. This also eliminates the need for insulated supply in the mixing stage from the baseband to the output of the mixing stage.

The output data stored in the send FIR filter 132 is subjected to correction for any error that may occur due to the 1/50 T difference between the I and Q time values. The IF filter in the RFU unit (Figures 28 and 29) also adds the two values ​​together to produce the correct transmitted wave signal, since its bandwidth is relatively small in relation to the intermediate frequency.

In the modem of the receiving portion, the mixing step 138 mixes an analog wave signal received from the RFU over the line 150 through a 20MHz bandpass filter not shown with a 20 MHz intermediate frequency signal arriving over the line 151 for downconversion of the analog signal to the baseband over the line 152. Then, the analog signal of the A / D converter 139 is converted to a digital signal over line 153 which is stored in the FIFO stack memory 140 for processing by microprocessor 141. Microprocessor 141 performs frequency and bit tracking on the received digital signal. , and it also performs FIR filtering and demodulation of the signal into a binary symbol stream which is passed over the line 154 to the CCU unit.

In addition to the analog and digital data signals processed by the modem, there are a number of control and status signals transmitted to and from the modem. These signals are commonly sent to the modem from the CCU. The modem also sends control signals to the RFU to control such functions as power output, frequency,

/...

124

AGC and antenna switching, in terms of diversity function.

FIG. 26 and 27 show the modem interfaces. The modem receives most signals from the CCU device. Other signals arrive from the RFU unit and the timing units. The input signals to the modem are as follows:

The following lines transmit the following signals from CCU 18, 29 to modem 19, 30a:

TX DATA lines 156 transmit a 4 bit symbol to be sent by the modem (2 bits for QPSK, 1 bit for BPSK). MOD BUS 157 is a bidirectional microprocessor bus that transmits control / status information to / from the modem. MOD WR line 158 transmits a control signal to latch MOD BUS into the modem. MOD RD line 159 transmits a control signal to place modem status and other information in MOD BUS for dispatch to CCU 18, 29. MOD RESET line 160 transmits a control signal to reset the modem. MOD ADD lines 161 transmit address signals to different locations for unlocking values ​​in the modem. The TX SOS line 162 transmits a signal to start broadcasting a TX interval. The RX SOS line 163 transmits a signal to start receiving the RX interval.

IF RECEIVE line 165 transmits an arriving intermediate frequency input signal from RFU 21, 31a to modem 19, 30a.

The following lines transmit the described signals from STIMU 35 to modem 19. The 80 MHz line 167 transmits an 80 MHz ECL clock signal. A timing unit not shown in the subscriber station provides the modem 30a with a corresponding signal. The 16 kHz line 168 transmits a TX CLK master signal used in the base station. The SOM line transmits a starting grid master signal from

STIMU in the base station. This signal is not used in the modem, but is passed on to the CCU 18, 29.

125

The following lines transmit the described signals from modem 19, 30a to the CCU unit 18, 29. The TX CLK line 171 transmits a 16 kHz clock signal providing the CCU unit with symbol timing. The symbols are read in the modem on the leading edge of the clock signal. In the base station, all intervals have the same TX CLK master signal. This will transmit all signals from the base station at the same time. At the subscriber station, the TX CLK signal of the modem is offset some time (fractional range delay) based on information from the CCU. The RX CLK line 172 transmits the 16 kHz clock signal derived from the received signal. This signal is always generated in the subscriber station, but is only provided in the base station during control interval capture. This clock signal reads the received symbol to the CCU unit, which also gets symbol timing. The RX DATA lines 173 transmit the received 4-bit symbol, clocked by the RX CLK signal. MOD BUS 157 transfers status and data information from the modem. MOD SOMF line

175 transmits the SOMF signal from STIMU to the CCU in the base station. AM STROBE line 176 transmits a shift from high to low to give the CCU a rough raster mark during the RCC capture at the subscriber station. It is a single pulse line that is activated when microprocessor 141 determines the approximate location of the AM hole.

The following lines transmit the described signals from modem 19, 30a to the individual RFU monitors 21, 31a. RF RX BUS 178 is an 8 bit bus between the modem and

RFU moiety. This bus transmits AGC information and frequency selection information to the RF RX section. The modem controls the AGC values ​​to be output and outputs CCU frequency selection information. The frequency selection information is passed from the CCU to the modem via MOD BUS 157. In training 35 mode, the modem controls the RF RX frequency selection. RF TX BUS 179 / ·

126 is an 8 bit bus between the modem and the RFU TX part. This bus transmits information about TX power level and frequency selection to the RFU TX section. The modem has nothing to do with these, so the information is only transmitted to the RF TX5 part. RX 80 MHz REF line 180 transmits an 80 MHz ECL reference clock signal to the RFU RX section. TX EN line 182 to the RFU The TX section transmits an RF unlock signal. RX EN line 183 to the RFU The RX section transmits an unlock signal for RF reception. AGC

WR line 184 transmits a write / strobe signal to input AGC data into the RFU RX portion. The RXFREQ WR line 185 transmits a write / strobe signal for frequency input into the RFU TX section. The PWR WR line 186 transmits a write / strobe signal to input power information into the RFU TX section. The PWR RD line 187 transmits a read / strobe signal to read power information from the RFU TX section. TXFREQ RD line 188 transmits a read / strobe signal to read back the transmit frequency from the RFU TX portion. The TXFREQ WR line 189 transmits a write / strobe signal for frequency input into the RFU TX portion. IF TRANSMIT line 190 transmits the signal transmitted at the intermediate frequency to the RFU array.

The following lines transmit the described signals from modem 19 to STIMU 35. VCXO BUS 192 is a 20 bit data bus with control information for frequency tracking to a voltage controlled oscillator in STIMU 35. VCXO WR line transmits a write pulse to the oscillator for unlocking from VCXO BUS 192 to the oscillator. Similar signals are transmitted from modem 30a to a timing unit not shown in the subscriber station.

The base station modem operation is assigned a fixed radio frequency. The communication at the base station is full duplex, which is why the modem's sending and receiving parts operate simultaneously. A modem is also assigned to the control frequency channel's modem, so it only transmits and receives

/.

127 information in RCC format during the allocated control interval period. All transmissions from the base station modems are clocked by the TX CLK master clock signal of 16 kHz over the line 171. Unlike the modem stations modems, the base station modems 19 to the CCU 18 deliver the fraction of the symbol time between the TX CLK master signal over the line 171 and the one derived in the modem 19 RX CLK signal over line 172. This information is then transmitted to the subscriber station over the RCC10 channel, so that the subscriber station postpones its transmission to ensure that its signal in the base station is received synchronously at all the other intervals.

The base station modem 19 also emits a zero energy signal in the control interval to create the RCC AM hole (which forms a raster reference) when the RFU emits a zero energy signal. This portion of the RCC transmission without carrier is used for original RX capture at the subscriber station.

Modem 19 does not realize that there are four voice codecs, multiplexed by CCU 18, in the base station for four 16 PSK subscriber interval assignments. Modem 19 accepts the bit stream from CCU 18 and treats the transmission as if there were only one codec subscriber.

All operations in the subscriber station modem

30a is derived from the RX CLK signal received over the line 172 which is extracted from the received transmission. This signal is used as the master rate signal for the subscriber30 station. The TX CLK signal over line 171 to CCU 29 is not a master clock signal as in the base station. It is derived from the RX CLK signal over line 172 and is imparted to the delay selected by CCU 29. CCU 29 determines the delay based on the radio control channel. The delay is determined by distance 175 1753 B1

128 the one between the base station and the subscriber stations. The subscriber station CCU 29 provides information about this delay to modem 30a through MOD BUS 157. Modem 30a even takes into account this delay.

CCU 29 takes into account the delay of the complete symbol by transmitting over the line 162 the TX SOS signal delayed by the correct number of symbols. In this process, the signals to the base station are matched, taking into account the distance to all subscriber stations.

In the subscriber station, communication is half-duplex. In this way, the transmitter is locked when idle. When not broadcasting, modem 30a is set to receive mode and thus able to monitor the power level of the received signal so as to be prepared when the base station calls.

The subscriber station modem 30a does not output any AM guard band for the RCC interval, as this is not required since it is the base station that determines the raster. Unlike the base station modems 19 which transmit at fixed frequency, the subscriber station modems 30a can also transmit or receive data over any of the 26 frequencies that the CCU 29 selects in the RFU.

There are many sources of delay in the modem that have a noticeable effect on system timing. These may be, for example, the delay of analogue filters, propagation delays, the processing delay of FIR filters, etc. These delays offset the TX and RX30 grids relative to each other and these delays must be carefully considered.

The delay between the TX SOS signal over line 162 at the base station and the tip of the first analogue symbol received at the base station is +7.4 sym35 boles. Therefore, there is displacement between TX and RX interDK 175353 B1

129 vallerne. In order to properly decode the arriving phase, the modem must start sampling approx. 3.5 symbols before the tip arrives. Therefore, the offset between the TX SOS signal and the beginning of an RX sampling is approx. 4 symbols.

at the base station, the onset of the RX interval occurs approx. 4 T after the start of the TX interval. The RX interval start is defined as the time when the first analog sample is taken for the purpose of detecting the first peak received.

The subscriber station clock signals are completely output from an 80 MHz master oscillator in the subscriber station timing unit not shown. The voltage controlled oscillator is controlled via an analog line from modem 30a. Based on this, all transmit and receive rate signals are calculated. Modem 30a supplies CCU 29 with the 16 kHz RX CLK signal over line 172, derived from the oncoming data stream. CCU 29 defects the unique word in the control channel itself and is able to determine the raster and interval markings from the unique word and RX CLK signal over line 172. The AM hole signal from the modem demodulated signal provides the CCU 29 message where it should look for the unique word.

During receipt of any interval, the modem 19, 30a performs frequency synchronization upon capture, after which it proceeds with detection. In the subscriber station, the voltage controlled oscillator is directly controlled by the microprocessor 141 through a D / aconverter. The frequency capture and tracking algorithms of the microprocessor calculate the necessary changes that the voltage controlled oscillator requires to maintain synchronization.

In the base station, the STIMU 35 includes a thermostatically controlled oscillator having a fixed frequency and

/.

130 determines the system master clock signal. Therefore, on the receiving side, there will be no frequency variation.

During receipt of any interval, modem 19, 30a also performs bit synchronization to the arriving data stream. An algorithm performs a bit-tracking loop at the receiver. Microprocessor 141 controls a variable frequency divider for the 80 MHz VCXO or OCXO oscillator (only during demodulation of control interval). Within the bit tracking loop, the microprocessor 10 changes the frequency sharing ratio to obtain bit synchronization. When receiving a voice channel, the split values ​​have step sizes of 0.1% of 16 kHz, while these values ​​change more drastically in a control interval, for example, as much as ± 50%.

Raster synchronization takes place quite differently in the base station and the subscriber stations. In the base station, the SOMP master signal is passed from the timing unit over line 169 to CCU 18 over line 175 via modem 19. This is the SOMP master signal used for all transmission from the base station. From this signal and the system's symbol master clock signal (16 kHz), the CCU can derive all interval and raster timing.

In the subscriber station, the raster synchronization of the CCU 29 is performed by detecting the unique word in the received RCC data stream. During the initial capture, modem 30a outputs a single pulse (AM STROBE) over line 176 for approximate raster marking. During capture, modem 30a searches for AM HOLE in the RCC channel. If AM HOLE is detected, rao30 demet 30a counts it into a few grids, whereupon, over line 176, it produces the AM STROBE marker signal for CCU 29 at the raster location of AM HOLE. CCU 29 uses this selection signal to initialize the raster marking counters (windowing), which CCU software can be modified for exact raster /.

131 synchronization. This also indicates that AM HOLE has been detected and that the RCC channel has been captured.

The interval synchronization is controlled by the CCU device

18, 29. Signals TX SOS over line 162 and RX

The SOS over line 163 is ordering the modem 19, 30a to begin transmission or reception of an interval. These signals are synchronized with the TX CLK signal over line 171 and RX CLK signal10 over line 172, respectively.

Self-training mode is a feedback mode in which the modem is brought in to train the coefficients of the receiver's digital FIR filter to compensate for the changes that the receiver's analog filters may make due to temporal and temperature changes. The analysis is done by returning the transmit data to the RF unit and receiving a known pattern in the receiver. The coefficients are optimized according to a Lagrange system on the basis of five following links: 1) the received data stream; 2) the data stream delayed by 0.05 T; 3) the data stream advanced 0.05 T; 4) the data stream from the adjacent upper channel and 5) the data stream from the adjacent lower channel.

In training mode, the microprocessor supplies 141 TX

The FIR filter 131 over line 143 with a series of training patterns of 32 symbols length. This is done through a FIFO stack storage that is not shown that is unlocked during this training mode. The advances / delays take place in the receiver bit-tracking circuit, displacing the two currents by 0.05 T.

The CCU 18, 29 sets the modem 19, 30a in training mode so that the modem's transmit portion can read special training data from the FIFO stack store in the modem. For some of these tests, the receiving portion is moved forward or back to 35. When the process is complete, the modem sends on

132

CCU unit 18, 29 a status message indicating that the coefficients are now calculated. At this point, the CCU 18, 29 tests the modem by switching it to normal operation and printing a given pattern on which

The RFU unit 21, 31a is switched back on, the data is sent back and checked for validity.

Such a modem is described in the parallel application DK 4270/85.

RF / RF unit (RFU) and antenna interface.

The RFU subsystem creates the connection between the modem and the antenna in both the base station and the subscriber station. The RFU operates linearly in terms of amplitude and frequency and is substantially transparent to the channel data and modulation.

FIG. 28 shows the antenna interface circuit for the subscriber station. An RFU control logic circuit 192 is coupled to the transmit antenna 32 and to the three antennas 32a, 32b and 32c through the antenna interface circuit. The logic circuit

192 is also connected to the transmitting portion of modem 30a and to the transmitting portion of modems 30a, 30b and 30c. In fact, the antennas 32 and 32a are the same antenna.

The transmit portion of the antenna interface comprises an upconverter and amplifier circuit 193, a TX synthesizer

194, a power amplifier 196 and a TX / RX switch

197th A first receiving portion RX 1 of the antenna interface comprises a down-converter and amplifier circuit 198, an RX synthesizer 199, and a preamplifier 200 coupled to switch 197. The individual additional diversity receiving parts TXn (n = 2, 3) include a downconverter and amplifier circuit 202, an RX synthesizer 203 and a preamplifier 204.

When it receives signals from the transmit portion of modem 30a, logic circuit 192 outputs the following signals to the transmit portion of the interface circuit: 1) a TX unlock /.

133 signal over line 206 to enable TX / RX switch 197 to unlock transmission from transmit antenna 32; 2) an IF input signal over line 207 to the up-converter and amplifier circuit 193; 3) a power control signal over line 208, also to circuit 193; 4) a clock reference signal over line 209 to TX synthesizer 194; and 5) a channel selection signal over line 210 to TX synthesizer 194. The TX synthesizer 194 responds to the signal over line 210 to transmit over line 211 to the upconverter and amplifier circuit 193 a TX frequency select signal equal to the difference between the desired transmit frequency and the modem's intermediate frequency.

In response to the signals received from the respective receiving portions of modems 30a, 30b and 30c, logic circuit 192 outputs the following signals to each of the receiving portions of the antenna interface circuit; 1) a TX unlock signal over line 213 to cause down-converter and amplifier circuits 198, 202 to operate in receive mode; 2) an AGC signal over line 214 to down-converter and amplifier circuits 198, 202; 3) a clock reference signal over line 215 to RX synthesizers 199, 203; and 4) a channel selection signal over line 216 to RX synthesizers 199, 203 depending on the channel selection signal across line 216, so that downconverter and amplifier circuits 198, 202 over line 217 receive a signal equal to the difference between the desired receive frequency and the modem. intermediate frequency. The down-converter and amplifier circuits 198, 202 output IF outputs over line 218 to logic circuit 192 for delivery to the receiving portion of the respective modems 30a, 30b and 30c.

The up-converter and amplifier circuit 193 in the receiving portion receives the modulated medium frequency signal over line 207, amplifying the signal and moving it.

/.

134 to the selected channel frequency. A combination of filters not shown, amplifiers 196, 197 and not shown power control circuits are used to create the appropriate output level and suppress unwanted signals on the mirror frequency and harmonic frequencies. The transmitted frequency is the sum of the modem's intermediate frequency and a converted frequency that is synthesized from the reference frequency of the modem in increments of 25 kHz.

The subscriber station RFU unit acts as a half-duplex transceiver when the receivers are inactive in the transmission intervals. The transmitted burst frequency is high enough to simulate full-duplex operation to the user. The assigned frequency channel is the one that the base station RPU unit selects.

FIG. 29 shows the antenna interface circuit for the base station. A control logic circuit 219 is coupled to the transmit antenna 23 and to three receive antennas 34a_z 34b and 34c through the interface circuit. The control logic circuit 219 also connects to the transmit portion of modem 19 and to the receiving portions of modems 19, 19b and 19c. Modems 19b and 19c are diversity modems not shown in FIG. 2nd

The transmit portion of the antenna interface circuit includes an up-converter and amplifier circuit 220, a TX synthesizer 221, a power amplifier 222, a high-power amplifier 223, a power detector 224 and a bandpass filter 225. A first receiving portion RX 1 of the antenna interface circuit includes a down-converter 230, an RX synthesizer 231, a preamplifier 232 and a bandpass filter 233. The individual additional diversity receiving portions RXn include a down-converter and amplifier 234, an RX synthesizer 235, a preamplifier 236 and a bandpass filter 237.

Depending on the signals received from the transmit portion of modem 19, the logic control circuit

135

219 the following signals to the transmit portion of the antenna interface circuit: 1) a TX ON signal over line 239 to the up-converter and amplifier 220 for activating the transmit portion for broadcast from the antenna 23; 2) an IF input signal over line 240 to the up-converter and amplifier 220; 3) a clock reference signal over line 24 of TX synthesizer 221; and 4) a channel selection signal over line 242 to TX synthesizer 221. TX synthesizer 221 responds to the signal over line 242 and outputs to upconverter and amplifier 220 over line 243 an RX frequency selection signal equal to the difference between the desired transmit frequency and the modem's intermediate frequency. The line 244 from the power detector 224 outputs a level control signal to the up-converter and amplifier 220.

In response to signals from the respective receiving portions of modems 19, 19b, 19c, logic circuit 219 outputs the following signals to each receiving portion of the antenna interface circuit: 1) an AGC signal over line 245 to down converter and amplifier circuit 230, 234; 2) a clock reference signal over line 246 to RX synthesizer 231, 235; and 3) a channel selection signal over line 247 to RX synthesizer 231, 235. The RX synthesizers 231, 235 respond to the signals over the wires 247 to output over the wires 248 to the down-converter and amplifier circuits 230, 234 an RX frequency selector signal equal to the difference between the desired receive frequency and the modem intermediate frequency. Above line 249, down-converter and amplifier circuits 230, 231 output intermediate frequency output signals to logic circuit 219 for delivery to the receiving portions of respective modems 19, 19b, 19c.

The RFUs in the base station and the subscriber stations are the same except for the additional high power amplifier 223 used to increase the transmit power

136 from the base station. The basic function of the RFU unit at both stations is to convert the modulated medium frequency signal (20.2 MHz) from the modem transmitter portion to the desired radio frequency in the UHF band (450 MHz). On the receive5 page, the RFU performs the reverse downconversion of the received UHF signal to a medium frequency signal of 20 MHz. The transmit and receive frequencies are offset by 5 MHz. The RFUs are programmed by the CCU to operate at the various frequencies used in the total system. Typically, each base station's RFU unit will be made to operate on a given frequency allocation when the system is initialized and this does not change. The number of RFU units in the base station corresponds to the number of transmit and receive frequency channels15 pairs that the base station can carry. The subscriber stations' RFUs will typically change the operating frequency for each new telephone call.

The RFU units are arranged so that the AGC level and the transmit power can be adjusted.

It is the modem that produces the AGC gain factor on the basis of a calculation in the transmitter portion processor 141. The subscriber station's transmit power is calculated by the CCU based on the messages received from the base station over the RCC channel as well as other control parameters.

If not all intervals in a frequency channel are used, the RFU will output a blank pattern that the CCU will set up in the channel. If a completed channel is not used, CCU software can interrupt the transmitter for this frequency through the modem.

The switching time for the diversity switches should be less than 50 us.

There are three antennas and three separate RF / IF units (single transmitter, three receivers).

Several parts of the base station RFU unit and anten35 neinterface are identical to the parts described above

137 of the subscriber station. The differences are explained in subsequent sections.

The base station RFUs and the antenna interface circuits operate full-duplex. All transmitters and receivers normally operate at 100% utilization ratio. Furthermore, it is economically advantageous to allow the base station to operate at a higher transmit power and to use low-noise and diversity-function receivers. The transmitter is designed to work with the maximum allowable power without dynamic control. Receive diversity is achieved by using multiple receiving antennas and multiple modems.

The base station does not normally change the operating frequency or transmit power during normal operation. The transmit15 and the receiving parts are fully tunable on each of the 26 channels.

The transmit portion of the base station antenna interface receives the modulated IF input signal over line 239 from the modem, and it processes this signal in the same manner as in the transmit station described above in the subscriber station. Furthermore, the signal is amplified to the required power and filtered in a bandpass filter 225 (so-called cavity preselector bandpass filter) to reduce the noise at the operating frequencies of adjacent receivers and to reduce the level of random transmitters.

The receiving portion of the base station antenna interface is designed as already mentioned for the subscriber station except that the antennas are coupled to bandpass filters

233, 237 (cavity preselector bandpass filters) which serve to prevent the receiver from being deafened by nearby or nearby transmitters. Low noise preamplifiers are also used to reduce the usable threshold signal level. There is 30 dB of isolation for all antennas 23, 34a, 34b, 34c opposite any one¹³⁸ preferably any other antenna. Further isolation is created in the sending and receiving parts to obtain approx. 80 dB of isolation between the transmitted signals and the received signals. The bandpass filter, preamplifiers and amplifiers are located in the immediate vicinity of the respective transmit / receive antennas.

Reception Diversity feature.

The Diversity function is used to reduce the sand10 visibility of channel fading below an accepted threshold. The Diversity system is capable of adding three-way diversity to both paths from base station to subscriber station and back. Diversity hardware in both the base station and the subscriber stations includes a special diversity combination circuit, three modems, the associated radio frequency units and antennas. There is only a single combination of modem, RFU device and antenna capable of broadcasting. Although the diversity combination circuit 33 is shown only for the subscriber station in the diagram of FIG. 2, it is present and coupled to the modems and the CCU unit in the base station in the same way as in the subscriber station.

When receiving using the diversity function, the base station or the subscriber station uses three receiving antennas that are sufficiently spaced apart that there is no correlation in the fading of the received signals. These three antennas through three identical receiving portions of the antenna interface deliver their signal to the RFU unit's control loop, whose IF output signal is passed to separate modems for demodulation. A microprocessor of the type TMS 320 in the combination circuit 33 (so-called diversity processor) receives the output signals from the modems and produces a more reliable data flow to the rest of the system in a way that emulates a single

/.

139 modem. The diversity processor's hardware and software are designed to perform the diversity combination and act as a single modem to the CCU.

The diversity processor reads from the three modems their data symbols, AGC values, signal + noise, size and phase error (deviation of the detected phase from the ideal 22.5 ° reference vectors). The algorithm used to determine the demodulated symbol involves the use of majority voting and calculations of signal-to-noise ratios for each modem in order to find the modem most suitable for providing the correct response signal .

The registers in the CCU interface of the diversity processor are almost identical to the registers in the modems except that the additional registers used to transfer information used in the diversity processing function are not needed, so only three address bits are needed.

Since the TMS 320 microprocessor's I / O capabilities are small and most of the processing works with one type of I / O register at a time, a special register is used containing the required register address at that time. For example, the AGC value from each modem must be read, the highest value selected, and the result entered into the diversity processor's I / O registers from which the CCU can read. Addressing these registers is most effective if the address of the AGC register is first loaded into a port where it is placed on the modem's address lines. Then the processor only needs to address that modem or microprocessor's register bank, which speeds up the I / O operations.

In the subscriber station diversity system, each modem has its own timing unit, and the timing signals used by the three modems in the diversity system are not

140 necessarily in phase. Since the modem clock signals for the three modems are not synchronized with each other, latch circuits are needed to keep it from each modem output symbol until the diversity processor reads this symbol.

An important function of the diversity processor is to maintain communications between the CCU and the three modems. This communication must be fast enough to meet all CCU requirements, but not so fast as to overload the diversity processor.

141