Method and apparatus for compressing and transmitting high speed data

Method and apparatus for high speed data compression and transmission

FIELD OF THE INVENTION

[0001] This invention relates to a communication unit and method and, more particularly, to signal processing techniques for compression of high-speed data communication signals for better transmission performance and greater capacity for transmission. System communication

BACKGROUND OF THE INVENTION

[0002] Telecommunications systems are well known in the art, and current telephone systems employ several multiplexing techniques to transmit telephone signals from many users on a single transmission line, such as a fiber optic cable or wire. Most of these systems with "direct wiring" employ a form of Time Division Multiplexing (TDM) in which multiple channels are transmitted sequentially at speeds greater than the information rate of the channel.

[0003] Typical telephone multiplexing requires sampling the telephone signal and transmitting the samples with a frequency much higher than the frequency of the telephone signal. To this end, current digital systems sample and encode the telephone signal, multiplex and transmit the signal, and then receive, demultiplex and decode the signal. One of these sampling and coding systems is Pulse Code Modulation (PCM) in which analog voice-band signals are sampled at a rate of 8 thousand samples per second in each sample represented by 8 bits. Consequently, the vocal band signal is converted to a digital signal of 64 kilobits per second (kb / s).

[0004] Another type of telecommunications system is the radio telephone system. Radio telephone systems use a group of radio frequencies (RF) selected to carry the telephone communication signals between two or more locations, and generally use a form of multiple frequency division access (FDMA). These radio systems, called wireless communication systems, are used, for example, in rural areas to provide local telephone services or in mobile units to provide mobile communication services.

[0005] One of the categories of RF communication systems employs TDM to allow users access to multiple time segments of the information modulated in the RF carrier. If many users compete for a small group of information time segments, the system is called multiple access by time distribution (TOMA). To allow the TOMA of FDMA RF communication channels, a method, called FDMAlTDMA, described in US Patent 4,675,863, has been used to increase the capacity of RF communication systems. However, RF communication systems continue to have their capacity often limited compared to communication systems with direct wire or fiber optic cabling.

[0006] Accordingly, to increase the capacity even further, signal compression techniques have been used to reduce the bandwidth necessary for the transmission of a telephone signal on an RF channel. The normal techniques used for voice signals are subband coding, adaptive differential pulse code modulation (ADPCM) and linear predictive coding (RELP). The RELP or similar voice compression algorithms allow transmitting a quantized and sampled voice signal of 64 kilobits per second (kbls) over the RF channel, such as a reduced bit rate signal (for example, 14.6 kbls or less). The receiver reconstructs the 64 kbls voice signal of the reduced bit rate signal, and the listener perceives little or no loss of signal quality.

[0007] The underlying method of voice compression, including RELP, is an encoding and decoding algorithm that takes advantage of the known characteristics of voice signals. One type of RELP method assumes certain characteristics of the harmonics of the human voice. Today, however, a large part of the communication signals in a telephone network are non-voice data communications signals, such as voice or facsimile (fax) modem data. Unfortunately, voice compression algorithms are not especially compatible with these data communications signals because the data signals do not have the characteristics of the voice signals.

[0008] Consequently, some RF communication systems monitor the telephone signal to detect the presence of a data communication signal. Normally, data signals have been detected that resend voice modem or fax band data signals of up to 2.4 kb / s (low speed data) and have been provided with a specialized compression algorithm. The receiver reconstructs the data signal, without reducing the speed of data transmission. This system and method is described, for example, in US Patent 4,974,099.

The current data telephone signals, however, are rather 9.6 kbJs (high speed data) or higher (ultrafast speed data, such as 14.4 kb / s 028.8 kbJs, or others, higher or lower), and current compression techniques do not compress these higher data rates satisfactorily. The compression of these higher data rates, and especially the multiple encodings of these

5 higher data rates, cause a degradation of the modem or fax signal quality, and the modem or fax machine will frequently reduce the speed of data transmission when the signals are passed through an RF communication system.

[0009J EP 0507384 A2 describes a wireless communication unit comprising means for transmitting data to a second communication unit using at least one of a plurality of data rates. at least one of a plurality of time segments, at least one of a plurality of modulation schemes, and at least one of a plurality of coding rates in a plurality of multiple access frames per time distribution. In US 5,533,004 A. A communication unit with this general structure is also described. Basically. EP 0507384 A2 describes a signal transmission system in which a carrier is divided into a plurality of time segments of the traffic signal. The durations of the time segments of the traffic signal in a frame may be the same or different to be able to manage transmissions between complementary users at different symbol rates and / or different modulation schemes having different numbers of levels. In particular, the length of a time segment is double for the low speed structure with respect to the high speed structure. Further. In order to alter the number of users that can access the system, the symbol rate remains the same, but the number of modulation levels is changed. Basically. US 5,533,004 A describes a radio frequency communication system that uses a TOMA to investigate multiple services, such as voice data and data packets, within the same RF communication channel. Specifically. Different data rate options are used to optimize the net speed of the amount of data that can be transmitted per unit of time of the user that will vary according to the RF characteristics. Likewise. different techniques can be chosen

25 of modulation according to the measures of estimation of the quality of the channel or the amount of user data to be transmitted or the destination of the data packet.

[0010J] It is an object of the invention to provide an improved communication unit and method for use by said unit, which has a higher performance with respect to the efficient use of the communication channel.

SUMMARY EXPLANATION OF THE INVENTION

[OOllJ A telecommunications system receives a group of telephone signals, including data signals each having a form of collation, and transmits the telephone signals in at least one radio frequency (RF) carrier. Each RF carrier has a group of information segments, and each telephone signal is assigned at least one information segment, so that the telephone signal is modulated in the RF carrier. The system includes a process for monitoring and identifying data signals, and for the

35 Compression of each data signal to reduce the necessary transmission bandwidth of the data signal.

BRIEF DESCRIPTION OF THE DRAWINGS

[0012J] The invention will be better understood by reading the following detailed description referring to the attached drawings, in which:

Figure 1 is a block diagram of a wireless communication system.

Figure 2 is a high-level block diagram of the application of the compression system of the present invention, including the dynamic bandwidth allocation function and the high speed and ultrafast speed data codecs.

Figure 3A is a high-level flow chart illustrating the detection and selection of high-speed data coding types and the determination and assignment of radio channel segments, according to a form of

Four. Five illustrative embodiment of the present invention.

Figure 3B is a high level flow chart showing the channel assignment process performed by the channel formation processor upon request for a high speed data channel, according to an embodiment of the present invention.

Figure 4A is a graph showing the characteristics of the quantifier of law A.

Figure 4B is a graph showing the signal for the execution of the PCM noise quantification versus the uniform quantification.

Figure 4C illustrates the compression method by assigning signal samples from one quantification to another quantification.

Figure 5A is a high level block diagram of the high speed data encoder, according to an illustrative embodiment of the present invention.

Figure 58 illustrates an encoding process for transmission with a high speed data encoder, according to an illustrative embodiment of the present invention.

Figure 6A is a high level block diagram of the high speed data decoder, according to an illustrative embodiment of the present invention.

Figure 68 illustrates a decoding process for transmission with a high speed data decoder, according to an illustrative embodiment of the present invention.

Figure 7A is a high level block diagram of the ultrafast speed data encoder, according to an illustrative embodiment of the present invention.

Figure 78 illustrates an encoding process for transmission with an ultrafast speed data encoder, according to an illustrative embodiment of the present invention.

Figure 8A is a high level block diagram of the ultrafast speed data decoder, according to an illustrative embodiment of the present invention.

Figure 88 illustrates a decoding process for transmission with an ultrafast speed data decoder, according to an illustrative embodiment of the present invention.

Figure 9 is a high level flow chart illustrating an ultrafast speed quantization algorithm used to allocate quantified PCM samples in compressed quantized samples, according to an illustrative embodiment of the present invention.

Introduction

[00131 A telecommunications apparatus and method receives the telephone signals and modulates each of the telephone signals in a corresponding radio frequency (RF) carrier. Each transmission RF carrier has a predetermined number of Information segments, and each telephone signal is assigned at least one information segment, so that the telephone signal is modulated in the RF carrier. The telecommunication apparatus and method induces a detector to receive and follow each of the telephone signals to detect a data signal contained in one of the telephone signals; and an encoder for encoding the data signal into a compressed encoded signal. The apparatus and method also induce a controller that controls an allocation status of each information segment when the data signal is detected and locates a predetermined number of sequential information segments not assigned (although not necessarily contiguous) for a bandwidth default required to transmit the compressed encoded signal. The assignment status indicates whether each segment of information has not been assigned or has been assigned to signals from another phone. The apparatus and method also induce a process to form a telecommunications channel from the localized unallocated sequential information segments and a process to modulate the encoded signal in the telecommunications channel.

[0014] According to one aspect of the present invention, a high-speed data compression transmission system transmits a high-speed data signal through a telecommunications channel such as a compressed coded signal. The high speed data signal is received as at least one block of sample data signals, and the system induces a high speed data encoder and a high speed data decoder. The high-speed data encoder induces: 1) a receiver for the blocks of data signals each containing at least one sample of the data signal representing a peak amplitude, 2) a calculator for calculating a gain value of the block of data signals, which is proportional to the value of the peak amplitude; and 3) a quantifier selector that selects a quantifier corresponding to the value of the gain.

[00151 The quantifier has a plurality of quantitative level values ​​with a given space (eg uniform), which are determined from the gain value, and the selected quantifier quantifies each data sample of the data signal block in A compressed data sample. The gain value and the plurality of compressed data samples constitute the compressed encoded signal. The high-speed data compression transmission system includes a transmitter to transmit the compressed encoded signal, through the telecommunications channel and a receiver to receive the signal from the telecommunications channel.

[0016] The high-speed data decoder of the high-speed data compression transmission system includes: 1) a receiver for compressed data samples and the corresponding gain value, and 2) a reverse quantizer selector to select , based on the value of the gain, a uniform inverse quantizer that has a plurality of uniformly spaced output values ​​that are determined from the value of the gain. The inverse quantizer processes each of the compressed data samples based on the gain value to provide a block of samples of the reconstructed data signal.

[0017] According to another aspect of the present invention, an ultrafast speed data compression transmission system of the system transmits an ultrafast speed data signal through a telecommunications channel. The ultrafast speed data signal is received as at least one block of sample data signals having a first quantification, and the system includes an ultrafast speed data encoder and an ultrafast speed data decoder. The ultrafast speed data encoder includes: 1) a receiver for the data signal block containing at least one sample of the data signal having a peak amplitude, 2) a calculator to calculate a gain value of the data signal block that is proportional to the peak amplitude, and 3) a quantifier selector to select a new set of quantification levels corresponding to the value of the sample block gain, and each of the new set of quantification levels are selected levels of the first quantification, and 4) a quantization level allocation processor that assigns the value of the signal from the signal to a value of the compressed level for each value of the signal sample based on a relationship between the set of levels of the first quantification and the new set of quantifying levels.

[0018] The gain value and the compressed data samples constitute a collated signal. The system also includes a transmitter to transmit the encoded signal through the telecommunications channel, and a receiver to receive the encoded signal from the telecommunications channel. The illustrative embodiment is described below with reference to a telecommunications channel of a wireless communication system. However, the present invention is not limited to wireless or other types of RF carrier communication. Rather, the present invention can also be used with the telecommunication channels of the cable communication systems to increase capacity.

[0019] The ultrafast speed data decoder of the ultrafast speed compression transmission system includes 1) a receiver for compressed data samples and the corresponding gain value, 2) a reverse quantizer selector to select, based on to the corresponding gain value, an inverse quantifier that has output values ​​that are determined from the value of the gain with the corresponding new set of quantizer levels. The inverse quantizer processes each of the compressed data samples based on the gain value to provide a block of samples of the reconstructed data signal.

[0020] According to another aspect of the present invention, a method for quantifying ultrafast speed data assigns, from a first plurality of samples of the quantized signal, each signal sample having its corresponding quantized amplitude value, and at least one signal sample having a peak quantized amplitude value, a second plurality of quantified compressed samples and a gain value. The method includes: 1) examining each amplitude to determine a peak amplitude value and establish the value of the gain corresponding to the peak amplitude value, and define for the first plurality of samples of the quantized signal a predetermined number of successive segments, each segment having a number of quantified level values. The quantified level values ​​for each successive segment are related to the gain value, and a first segment of the predetermined number of successive segments corresponds to the peak amplitude of the plurality of signal samples.

[0021] The quantification method also includes the allocation of each of the samples of the quantized signal in quantified samples compressed by 1) the retention, for each of the values ​​of the quantized signal, of those selected from the number of Quantified level values ​​for each segment until a zero value level is found and 2) the establishment of a sign value at a negative value to indicate a negatively valued amplitude.

Detailed description of the invention

The data compression system

[0022] Figure 1 is a diagram of a wireless telecommunication system in which the characteristics of high speed data compression of the present invention can be implemented. As can be seen, the radio telecommunication system includes a base station 11 and a group of subscriber units 10. The base station 11 communicates simultaneously with the subscriber units 10 by broadcasting and receiving defined communication channels in a range of preselected radio frequencies. The base station 11 can also interconnect with the local telephone equipment in the central office of the telecommunications company 12.

[0023] A typical radio telecommunication system (for example, the SLS-104, manufactured by InterOigítal Communications Corporation, King of Prussia, Pennsylvania) uses 24 direct predefined channels (from the base station to the subscriber unit) and 24 inverse predefined channels (from the subscriber unit to the base station) in the spectral region of 300-500 Megahertz (MHz). Communication from the base station to the subscriber unit is provided through pairs of communication channels (direct and reverse) modulated in frequencies within this spectral region. In a typical system. base station 11 communicates simultaneously through these 24 pairs of channels. The 24 channels can occupy. for example, 2 MHz frequency bands. The 2 MHz frequency band can support more channels. for example, 80 channels. using a 25 kHz channel separation. In a tonna of realization of the system. the base station 11 can transmit to a subscriber in the lowest frequency of a pair, and the subscriber unit 10 can transmit to the base station in the highest frequency pair. This system is described in US Patent 4,675,863. granted on June 23, 1987. titled TELEPHONE RF PAYMENT SYSTEM TO PROVIDE MULTIPLE VOICE SIGNS AND / OR DATA SIMULTANEOUSLY IN ONE OR A PLURALITY OF RF CHANNELS. on behalf of Paneth et al.

[0024] In order to increase communication capacity. Multiple access techniques are used by time distribution on each carrier frequency. In an illustrative system. each frequency of the channel pair is divided into four time segments such that the base station 11 communicates simultaneously with up to four subscriber units 10 on a carrier frequency. In consecuense. The base station, which uses 24 pairs of channels, can allow the telephone signals to be modulated in 95 channels, and use a channel for control and other general functions.

[0025] One aspect of increasing the capacity of this source is to compress the telecommunication channels that are transmitted through the RF communication channel (or the cable channel). For the voice as described above. Voice coding techniques can be used. such as RELP. Further. Low speed data compression and low speed fax data can be used. as described in US Patent 4,974,099 entitled COMMUNICATION SIGNAL SYSTEM AND METHOD OF COMPRESSION, in the name of Un et al.

[0026] In the system described above. three vocal band encoders, RELP. Low speed data and low speed fax, compress PCM signals at 64 kbls in a signal at 14.5 kb / s. At 14.5 kb / s. These three encoders can operate within a single 16-phase RF segment or a double-wide 4-phase RF segment. The RELP encoder is used for voice. The low speed data encoder is used to pass a number of voiceband modem transmissions at speeds of up to 2400 bps. and the low speed fax encoder is used to pass Group 3 fax transmissions at 2400 bps. Each transmission encoder has a corresponding decoder within a receiver. That can, for example. be assigned through the system control channel.

[0027] To allow the telecommunication system to adapt to high-speed voice band fax machines and modems. The two related vocal band compression techniques of the present invention are employed. Encoders and decoders (codecs), called high speed codec and ultrafast speed codec. they achieve a better execution of compressed data transmission than low speed and fax data encoders. employing less compression and therefore providing more bandwidth to the data signal.

[0028] The high-speed codex allows the telecommunications system to pass fax and modem transmissions in voice band at speeds of up to 9.6 kb / s. The ultra-fast speed code supports fax and modem transmissions in voice band up to 14.4 kb / s. and higher. The high-speed code works using three 16-phase RF segments or four 8-phase RF segments. The ultrafast speed codec works using four 16-phase RF segments. Preferably. High-speed and ultra-fast data compression algorithms pass a representation of an analog vocal band waveform through a digital channel with limited data rates while minimizing perjUcial distortion.

[0029] Because these codecs use several RF segments. Dynamic reallocation of segments in RF communication channels is necessary. The dynamic time segment / bandwidth allocation feature of the present invention detects and tracks data transmission and forms a data channel from the necessary number of segments. but if the necessary number of segments were not available. The low speed data encoder or low speed fax data is assigned to the call. These allocation methods are described. for example. in US Patent 4,785,450. granted on November 15, 1988. entitled DEVICE AND METHOD FOR OBTAINING FREQUENCY AGILITY IN DIGITAL COMMUNICATION SYSTEMS, in the name of DR Bolgiano et al.

[0030] Figure 2 is a high-level block diagram of the application of the compression system of the present invention. including the dynamic allocation function of time / bandwidth segments, and high-speed and ultra-fast data codecs, for high-speed data compression of the illustrative embodiment of a wireless telecommunications system. The system includes: a compression selector processor (CSP) 200, which includes a control unit 201 and a tracking section 202, a channel formation process 260, and the compression encoders / decoders (codecs) RELP 210, data low speed 220, low speed fax 230, high speed data 240 and ultrafast speed data 250.

[0031] The CSP 200 receives the telephone signal from the telephone of the local switchboard 270 and is a digital processor designed to implement the monitoring of the telephone signal to identify the specific types of data signals with their respective modem response tones and start the establishment of the communication channel. In another illustrative embodiment that uses subscriber to subscriber communications, the CSP 200 can receive the telephone signal from other local sources. The tracking section 202 of the CSP 200 informs the control unit 201 of the presence of the data signal. The control unit 201 is responsible for implementing the external formation of an RF communication channel, as well as the assignment of a type of compression CODEC 210, 220, 230, 240 and 250.

[0032J] The channel formation processor 260 receives a transmission channel request from the CSP 200 and assigns an available RF communication segment to a telephone signal. The channel formation processor 260 maintains the allocation information of the current system channel in a memory (not shown) to determine which time segments are not currently used for other telephone signals. As is known in TOMA systems, each time segment of the channel is formed with a guard time, which is a short period of signal used to initialize a receiver before sending the data. If there are data signals that require more than one RF time segment, the channel formation processor 260 forms the channel from a predetermined number of time segments, and if the predetermined number of time segments is contiguous, only Guard time is used.

[0033J] The channel forming process 260 of an exemplary embodiment of the invention may be a radio processing unit (RPU) of a network base station. The RPU may be responsible for storing channel time slot assignments and allocating channel time segments for the entire system of Figure 1.

[0034J CODEC RELP 210 implements compression coding (and decoding) algorithms for voice signals. The low speed data CODEC 220 and the low speed fax CODEC 230, the high speed data CODEC 240 and the ultrafast speed data CODEC 250 implement the respective data compression algorithms for the voice band data of the identified type

[0035] In general, CSP 200 and CODEC 210, 220, 230, 240, and 250 can be integrated into a digital signal processor to implement data signal tracking, signal processing and coding operations and decoding of signal compression. Such a processor is chosen, for example, from the family of TMS 320C5X digital signal processors from Texas Instruments.

[0036] The operation of the compression system of the present invention is described below. Referring still to Figure 2, when the voice call is first established, voice RELP codec 210 is initially assigned to the telephone signal. The CSP 200 tracks the telephone signal through the tracking section 202 and the control unit 201 determines the type of vocal band signal based on the detection of the modem response signal. Each type of vocal band data has a particular and identifiable modem response signal. Table 1 summarizes some of the different typical characteristics of the origin and response of the modem that are well known in the art. Table 1 serves illustrative purposes and is not intended, however, to describe all possible features of the modem.

TABLE 1

<dl><dt>Answer (or return channel) </dt><dd>Origin </dd></dl>

<dl><dt>V.77 / BPS </dt><dd>Disable EC Duplex you mod fa brand space faith mod fa brand space </dd></dl>

<dl><dt>V.16 </dt><dd>NO B 480 FSK 100 570 390 950 1400 2100 FM </dd></dl>

<dl><dt>V.16 digital </dt><dd>NO B 480 FSK 200 570 390 1400 FSK 100 1480 1320 </dd></dl>

<dl><dt>V.19 </dt><dd>NO 420 A.M 5 OTMF </dd></dl>

<dl><dt>V.19a111 </dt><dd>NO B 1750 FSK &lt;= 300 1850 1650 OTMF </dd></dl>

<dl><dt>V.19 a112 </dt><dd>NO B 420 FSK &lt;= 75 390 450 OTMF </dd></dl>

<dl><dt>V.20 </dt><dd>NO B 420 A.M 5 MTFSK 920-1960 </dd></dl>

<dl><dt>V.20alt </dt><dd>NO B 460 FSK &lt;= 75 420 480 MTFSK 1960 ~ 20-1960 </dd></dl>

<dl><dt>V.21 </dt><dd>2100 F 1750 FSK &lt;= 300 1850 1650 1080 FSK &lt;= 300 1180 980 </dd></dl>

<dl><dt>V.2212oo </dt><dd>2100 F 2400 40PSK 600 1200 40PSK 600 I </dd></dl>

<dl><dt>V.22 bis 2400 </dt><dd>2100 F 2400 16QAM 600 1200 16QAM 600 </dd></dl>

<dl><dt>V21.12oo </dt><dd>2100 B 420 FSK &lt;= 75 390 450 1700 FSK &lt;= 1200 1300 2100 </dd></dl>

<dl><dt>V.23600 </dt><dd>2100 B 420 FSK &lt;= 75 390 450 1500 FSK &lt;= 600 1300 1700 </dd></dl>

<dl><dt>V.26b </dt><dd>2100 FIH 1800 40PSK &lt;= 75 1800 40PSK 1200 </dd></dl>

<dl><dt>V.27 ter 4800 </dt><dd>2100 H 1800 80PSK 1200 1800 80PSK 1200 </dd></dl>

<dl><dt>V.27 ter 2400 </dt><dd>2100 H (B) 1800 40PSK 1200 1800 40PSK 1200 </dd></dl>

<dl><dt>V.299600 </dt><dd>2100 H 1700 16QAM 2400 1700 16QAM 2400 </dd></dl>

<dl><dt>V.2972OO </dt><dd>2100 H 1700 8QAM 2400 1700 8QAM 2400 </dd></dl>

<dl><dt>V.2948oo </dt><dd>2100 H 1700 4PSK 2400 1700 4QAM 2400 </dd></dl>

<dl><dt>V.329600 </dt><dd>2100 F 1800 16QAM 2400 1800 16QAM 2400 </dd></dl>

<dl><dt>V.3248OO </dt><dd>2100 F 1800 4PSK 2400 1800 4PSK 2400 </dd></dl>

<dl><dt>FAX channel 300 OPS </dt><dd>H 1800 FSK - ~ ------_.</dd></dl>

Local band modem features

[0037] Returning to Figure 2, once the type of vocal band data is determined, if compression of high speed data or ultrafast speed data is required, the CSP 200 initiates the reassignment of the voice channel, and The method of dynamic allocation of time segments used is described below. The control unit 201 instructs the channel formation processor 260 to form an RF communication channel with a predetermined number of time segments. In an embodiment of the present invention, a time segment is automatically assigned to the call, but this is not necessary. The channel formation processor 260 examines the memory to determine the number and location of the RF carrier of the available RF time segments. If the channel formation processor locates the number of segments

10 By default, the RF communication channel is formed from the predetermined number of RF time segments and the control unit 201 is notified. The control unit 201 then assigns a high-speed data codec or a data codec6. of ultrafast speed corresponding to the data signal, and the compressed data signal is assigned to and modulated in the multi-segment RF communication channel formed.

fifteen [0038] If there are not enough time segments available, the control unit 201 is informed and an RF communication channel is formed from a single RF time segment, and the control unit 201 then assigns the CODEC Low speed data or low speed fax CODEC to the data signal. As indicated above, an embodiment of the present invention automatically allocates a time segment when the telephone signal is received before forming a multi-segment communication channel of

twenty time, and therefore a segment has already been assigned to the telephone signal at this time.

Dynamic allocation of time segments / bandwidth

[0039] Table 2 summarizes the requirements of the time segments for the types of signal compression:

TABLE 2

<dl><dt>Encoder </dt><dd># 4 phase segments # 8 phase segments # 16 phase segments </dd></dl>

<dl><dt>RELP </dt><dd>two N / A 1 </dd></dl>

<dl><dt>Low speed data </dt><dd>two N / A 1 </dd></dl>

<dl><dt>Low speed fax </dt><dd>two N / A 1 </dd></dl>

<dl><dt>High speed data </dt><dd>N / A 4 3 </dd></dl>

<dl><dt>Fast speed data </dt><dd>N / A N / A 4 </dd></dl>

25 [0040] Since the high-speed encoder modulates the data in channels of three segments of 16 phases and four segments of 8 phases, its compressed data desirably fits into one of the two channels that has less bandwidth. Table 3 shows the availability of bits of the various types of channels of the arrangement for the radiotelecommunication system described in Figure 1.

TABLE 3

<dl><dt>Mod, Level, Segments </dt><dd>Mode start zeros Preamble CW block A block B ending zeros Data bits I Block </dd></dl>

<dl><dt>16-PSK, 1 </dt><dd>voice / test dean! OR 5 3 80 84 8 328 </dd></dl>

<dl><dt>16-PSK, 3 </dt><dd>voice (HSD) OR 5 3 262 262 8 1048 </dd></dl>

<dl><dt>16-PSK, 4 </dt><dd>voice (UHSD) OR 5 3 352 352 8 1408 </dd></dl>

<dl><dt>8-PSK, 1 </dt><dd>channel test OR 14 4 154 OR 8 462 </dd></dl>

<dl><dt>8-PSK, 4 </dt><dd>voice (HSD) OR 14 4 347 347 8 1041 </dd></dl>

<dl><dt>4-PSK, 2 </dt><dd>voice / channel test OR 13 6 160 173 8 328 </dd></dl>

<dl><dt>BPSK, 1 </dt><dd>RCC 8 44 8 112 OR 8 112 </dd></dl>

<dl><dt>HPSK, 1 </dt><dd>Refinement OR 52 8 (UWJ 112 OR 8 112 </dd></dl>

[0041] In Table 3, "zeros" indicates that there is no modulation, the preamble is a bit synchronization pattern, and "CW" means code word, which includes call control, call processing and signaling information Block A and Block B represent a first and second block of 22.5 ms of compressed vocal band data samples.

[0042] As seen in Table 3, the 8-phase four-segment channel carries fewer bits than the 16-phase three-segment channel. The compressed output block of the high-speed encoder of an embodiment of the present invention, therefore, can occupy 1041 bits or less. Table 4A shows the

10 Bit allocation of the compressed output block of the high-speed data encoder.

TABLE 4A

<dl><dt>Data </dt><dd>Bits per Instance Quantity Protected Number of bits </dd></dl>

<dl><dt>Coded sample </dt><dd>5 180 yes 900 </dd></dl>

<dl><dt>Coded Profit </dt><dd>6 1 yes 6 </dd></dl>

<dl><dt>Protected reserve </dt><dd>1 6 yes 6 </dd></dl>

<dl><dt>Hamming parity </dt><dd>7 16 N / A 112 </dd></dl>

<dl><dt>Reservation </dt><dd>1 24 no 24 </dd></dl>

<dl><dt>Total per block </dt><dd> 1048 </dd></dl>

[0043] In Table 4A "protected" indicates that direct error correction (FEC) is applied to the bit stream. The bitstream of the ultrafast speed encoder modulates a 16-phase four-segment channel, of which

1,408 bits are available for encoder data in each time period of 22.5 ms.

[0044] Table 48 shows the bit allocation of the compressed output block of the ultrafast speed data encoder.

TA8LA48

<dl><dt>Data </dt><dd>Bits per Instance Quantity Protected Number of bits </dd></dl>

<dl><dt>Coded sample </dt><dd>7 180 yes 1260 </dd></dl>

<dl><dt>Coded Profit </dt><dd>7 1 yes 7 </dd></dl>

<dl><dt>Protected reserve </dt><dd>13 1 yes 13 </dd></dl>

<dl><dt>Hamming parity </dt><dd>7 16 N / A 112 </dd></dl>

<dl><dt>Unprotected reserve </dt><dd>16 1 no 16 </dd></dl>

<dl><dt>Total per block </dt><dd> 1408 </dd></dl>

[0045] The high speed data compression and ultrafast speed data techniques described in

5 Below are embodiments of the present invention that may require multiple time segments for a communication channel, although other compression techniques may be developed in the same spirit as described herein for other specific types of data signals that do not necessarily follow the vocal band modem features described above. These other embodiments may also employ the method of dynamic allocation of time / bandwidth segments that is used in the

10 present invention

[00461 The general method of dynamic allocation of time / bandwidth segments is described below. Figure 3 illustrates the process of dynamic allocation of time / bandwidth segments as applied in, for example, the CSP 200 of Figure 2. Referring to Figure 3A, when the voice call is first established Once, the voice tracking stage 301 tracks the phone to detect a

fifteen data signal In step 301, the RELP c6dec 210 is initially assigned to the telephone signal. However, when there is a data signal, decision step 302 determines the type of vocal band signal based on the detection of the modem response signal.

[0047] If the data is low speed data or low speed fax, step 303 assigns the low speed assignment process to which, for example, a single RF carrier segment has been assigned. TO

twenty Next, step 304 determines whether the data signal is FAX or low speed data, and allocates the respective algorithm stages 305 and 306 of the low speed fax codec 230 or of the low speed data codec 220.

[0048] If the signal is of a high speed data type in step 302, then, the next step 307 requests a high speed data channel of the channel formation process 260. In an embodiment of the

25 In the present invention, the channel formation process 260 will require the user / subscriber to provide information to request the type of channel. Another embodiment of the present invention can also determine, from the modem signals, whether the data signal requires the method of compression of high speed data or ultrafast speed data in order to request the correct type of channel .

[0049] Figure 38 shows the channel assignment process performed by the channel formation processor

30 260 after requesting a high speed data channel from step 307 of Figure 3A. The channel-forming processor may be a radio processing unit (RPU) of the base station of the prior art system described above, and the RPU may allocate RF carrier time segments to communications of subscribers through a communication channel.

[0050] Starting at step 320 of Figure 38, the processor normally assigns a voice channel for a

35 phone call; however, any initial assignment process can be chosen, as described in US Patent 4,675,863. Next, step 321 checks if there is a request for a high speed data channel from step 307 of Figure 3A. If there is no request, the assignment is maintained in the default mode, which is voice for this illustrative embodiment. If there is a request, step 322 controls the provisioning of the subscriber to determine if the subscriber is ready to accept a data channel of

40 high speed. If the subscriber is not prepared to accept a high speed data channel, a low speed fax / data channel is assigned in step 323 using a predetermined number of segments.

[0051] If the subscriber is prepared for a high speed data channel, step 324 determines whether the subscriber is prepared to accept a high speed data channel of the ultrafast speed type ("UHSD Channel") (or if Has been requested). If so, step 325 checks if there is a predetermined number of RF carrier segments 45 available, and if so, step 326 creates the UHSD Channel. Stage 325 can be performed by

a tester that controls its memory that contains the allocations of the current system channels to know IF a necessary number of sixteen phase RF time segments is available (four for the illustrative embodiment) . If the required number of segments is not available. then try to find out if the channel can be created as one of high speed data type ("HSD Channel") as described later in step 328.

[0052] If the subscriber's prevention (or the request) indicates that the high-speed data channel should not be formed as an ultra-fast speed channel of the UHSD channel type in step 324. step 327 checks whether the request

or subscriber prevention indicates that the high-speed data channel should be formed as a high-speed HSD type channel. Otherwise. the low speed data channel is formed in step 323. as described above. but if the HSD Channel is requested or provisioned. then step 328 verifies if the predetermined number of RF carrier time segments is available for the HSD channel.

[0053] Step 328 may be performed by a processor that checks whether a memory contains the allocations of the current system channels to find if a necessary first number of time segments (sixteen phase RF time segments) is available ( three for the illustrative embodiment). And if it's not that way. if a second necessary number of time segments (eight phase RF segments) is available (four for the illustrative embodiment). If the necessary number of time segments is available. the time segments are assigned and the HSD channel is formed in step 329. If the high speed channel availability stage cannot find the necessary number of channels. then step 323 simply allocates the low speed channel.

[0054] Returning to Figure 3A. in step 308. the process checks the response to the request for the high speed data channel. If the request is denied in step 308 and no high-speed data channel has been formed. then steps 303 and the sequence to assign the low speed algorithms are executed. If the request for high speed data channel is accepted. the high speed channel availability stage 309 determines what type of channel has been assigned. If the high speed data channel corresponds to the ultrafast speed data. the ultrafast speed data CODEC coding algorithms 250 are executed in step 310. and if the high speed data channel corresponds to the high speed data, the coding algorithms are executed in the high data CODEC speed 240 in step 311.

The high speed and ultrafast speed CODEC

[0055] High speed codec 240 and ultrafast speed codec 250 provide compression of a bidirectional data channel of the present invention with sampled telephone signals (Pulse Code Modulation (PCM ', the telephone signals being, in the Illustrative arrangement, the input signal and the output signal. The telephone signals provided to the compression process are usually of PCM of law A or law Mu of 64 kbls, although whole samples of 16 bits at 128 kb / s, or otherwise, can be used by using a conversion process . The compression process compresses the sample bit flow of 64 kb / s (or 128 kbs) at a lower data rate. The lower speed data is sent by the RF channel for the expansion process, which again expands the lower speed data in a bit stream of the reconstructed sample of 64 kb / s (or 128 kbls). The purpose of the encoder is that the synthesized or reconstituted samples be an approximate representation of the original sampled signal.

[0056] In PCM systems. Analog vocal band signals are converted into a sequence of digital samples at a sampling rate of 8 thousand samples / second. The samples have a width of 8 bits, which allows 256 possible quantification levels. When analog signals are sampled, an important factor of merit is the signal to quantification noise ratio (SQNR). For a uniformly separated quantifier, the SQNR is 6B-1.24 dB, where B is the number of bits per quantized sample.

[0057] A uniform 8-bit quantifier therefore has an SQNR of 46.76 dB, which is excellent for voice signals. This SQNR is only achieved if the original analog signal has an amplitude that occupies the entire dynamic range of the quantification. If the dynamic range of the original signal exceeds that of the quantifier, a limitation occurs. This is a very undesirable type of distortion for both voice and voice band modem signals. If the original signal has a dynamic range less than that of the quantifier, the resulting SQNR is less than the optimum of 46.76 dB. For every dB that the dynamic range of the signal is less than the dynamic range of the quantifier, there is a loss of 1 dB of the SQNR.

[0058] Since the voiceband signals used in telephony have a wide dynamic range, a uniform quantifier may not be the optimal option. Therefore, non-uniform quantifiers are not used. There are two standards for non-uniform quantifiers for PCM: The Mu law and the A law, and these standards are well known in the art and are described in Chapter 8, Communication Systems, by Siman Haykin. Both techniques use log levels separated logically in order to increase the dynamic range of the quantifiers. Figure 4A shows the characteristics of the quantifier of law A.

[0059] the separation between the levels of the quantifiers is greater than the separation at low levels. The result is a more uniform SQNR in a sample for sampling. Although the best SQNR for these quantifiers is less than that of the uniform 8-bit quantifier. These quantifiers can provide a good SQNR over a wider range of signal levels.

[0060J] Figure 48 compares the result of the SQNR with the signal level of a law A and a uniform 8-bit quantizer. Although the uniform quantifier shows superior performance at high signal levels. the quantifier of law A retains a good SQNR in a wider dynamic range.

[0061] Voiceband modems work well in a telephone network that uses any of the PCM Mu or A laws at 64 kbJs due to the wide dynamic range. The transmission output level of these modems is high in order to use the channels at maximum capacity. but telephone channels have various signal level losses. As a result. Although the modem output level is set to a high level, the level at another point in the network may be significantly lower. The dynamic range of the PCM compensates for this situation.

[0062] PCM compression of 64 kbJs at a lower data rate decreases the number of bits per sample and generally produces a significant decrease in the SQNR. The distortion produced by compression is minimized. Through the present invention. dynamically designing a quantifier to adjust the dynamic range of the input signal. Once the two dynamic ranges correspond, the samples are quantified using a quantifier with the separation of levels just defined.

[0063] Figure 4C illustrates a simple example of the compression method by assigning the signal samples from one quantification to another quantification. A block of signal samples 410 is composed of three samples 411, 413 and 415. A first quantification level group 420 indicates the approximate value of the amplitudes of the sample 412. 414 and 416. However. Quantification levels require a certain number of bits of information. Five bits for the 20 levels shown from the first quantification. they are transmitted to a receiver to represent one of the levels of the first quantification. To send three sample values ​​corresponding to the three samples 411, 413 and 415. Fifteen bits are desirable.

[0064] The illustrative method of the present invention defines a new set of levels for each block of signal samples based on peak amplitude. As shown in Figure 4C. Sample block 410 has a sample 413 that has a peak amplitude value 414. The method defines a new set of quantification levels by defining peak amplitude 414 as the highest level value. and determines a predetermined number of level values ​​below this amplitude. As shown in Figure 4C. This corresponds to 5 level values. For this new quantification. Only three bits are necessary to define a level value. but the peak amplitude value must also be sent as a scale factor to indicate the relationship between the new quantizer level values ​​and the original quantizer level values. In consecuense. or five bits corresponding to the original peak amplitude value and nine bits (three per sample) are transmitted for sample block 410, or fourteen bits are required. The example shows that one bit less is sent, however. If there are ten samples in the block, the original quantization method requires the sending of fifty bits. but the new quantifier only requires sending thirty-five bits.

[0065] Next, the embodiments designed for the standards of the Mu law and the A law are denied. However. The described techniques easily extend to any quantized sample reception system with a non-uniform compression and expansion quantifier.

The high speed data CODEC

[0066] Figure 5A is a block diagram of a high speed encoder. The encoder of the exemplary embodiment transforms the data between the 64 kbJ PCM and the forward encoded compressed data stream (FEC) of 46.58 kb / s. The compressed data rate is 40,267 kb / s. and the rest of the transmitted bit stream is used for error correction.

[0067] As shown in Figure 5. the high-speed data encoder of the present invention includes an optional buffer 510. a PCM expander 520. a gain calculation process 522. a delay 521, a sample quantifier of data 523. and an optional 530 transmission encoding process. The transmission coding process 530 further includes an FEC encoder 532 and an interleaver 531.

[0068] The optional buffer 510 has a predetermined number of samples to create a sample block for the high-speed data compression process. Thematically. Samples can be received in a block format. The PCM 510 expander converts the PCM samples from law A or law Mu into linear samples. The gain calculation process 522 calculates the value of the quantized gain for the sample block. and the quantifier of data samples uses the quantized gain value to create a quantifier uniformly separated with the quantization level values ​​graduated by the quantized gain value. B delay shows that the quantized gain value is determined before the compression process creates the encoded quantified samples and the transmission coding process 530

it is used to provide error correction coding for the transmission of the encoded quantized gain and the encoded quantified samples.

[0069] The operation of the high speed data compression encoder is described below. As shown in Figure 5A, the 64 kb / s PCM samples (law A or Mu law) are received by a buffer 510. Buffer 510 provides the PCM samples as blocks of 22.5 milliseconds of samples. At the rate of 8 thousand samples / second of the PCM, each block contains 180 samples. The frame of the received PCM is introduced into the PCM 520 expander, which converts the Mu law or A law samples into 16-bit linear samples (16-bit whole samples).

[0070] The resulting block of linear samples, which are 16-bit whole samples in the illustrative embodiment, is introduced in the gain calculation process 522, which finds the sample in the block with the largest amplitude value ( absolute value). The amplitude of this sample determines the value of the quantized gain for the block. The quantified gain value may be the amplitude value, the difference between the maximum value of the sample and the amplitude of the largest block, or a multiplier value. The value of the quantified gain is quantified using a logarithmically separated quantifier of 64 levels. The gain calculation process 522 provides both the quantized gain value and the encoded quantified gain value. The encoded quantized gain value is a 6-bit number that represents one of the 64 levels in the gain quantifier logarithmically separated.

[0071] The value of the quantized gain of the gain calculation 522 and the sample block of the PCM expansion process are provided to the quantizer of data samples 523. The delay 521 is shown to indicate that the gain calculation process 522 You must complete the task in the block before the samples are compressed by the 523 data sample quantizer. The data sample quantifier 523 quantifies the 180 samples in the block using a uniformly separated quantizer of 32 levels. The quantifier levels are dynamically adjusted block by block using the value of the quantized gain. Therefore, the levels of the uniformly separated quantifier vary from + quantized gain value to -quantified gain value for the current set of 180 samples. The sample quantifier issues only the 5-bit coded representation of the 180 samples, as compression does not require the actual quantized values.

[0072] The encoded quantified gain and the encoded quantified samples are optionally introduced into the transmission encoding process 530, which includes interleaver 531 and FEC encoder 532. The FEC encoder 532 is an extended Hamming encoder (64.57) and The Hamming code is capable of correcting a one-bit error and detecting a double-bit error in each 64-bit block. The FEC encoder 532 receives the encoded quantized gain and the encoded quantified samples and provides them to interleaver 531, and interleaver 531 issues the encoded compressed data. The interleaver of an illustrative embodiment of the present invention is a 16 * 64 bit block interleaver.

[0073] Figure 58 shows an illustrative embodiment of the transmission coding process 530 including interleaver 531 and Hamming FEC 532 encoder. A block of 64 by 16 bits is shown. Each of the 16 rows represents a single 64-bit extended Hamming code word. In the encoder, the data in the interleaver block is read from left to right through the rows starting with the code word O bit OY ending with the code word 15 bit 63. The positions of the bits (columns) 0.1 , 2,4,8, 16 and Y32 are omitted and filled with zero. After filling interleaver 531, Hamming coding is performed by means of the FEC 532 encoder in the 57 data bits in each row. The Hamming parity bits are entered in the positions of bits 1, 2, 4, 8, 16 and 32, as shown in the diagram. The parity check bit is entered in the O bit position. The parity bits and the parity check bits for the 16 codes can be calculated at the same time using a wide 16-bit O-Exclusive tuning. The parity bits Pi are calculated as follows:

Pi = XOR Word code Bit [k] i = 0. • 6

(k-1) &amp; 2 '* O; where "&amp;" is a bit-by-bit ANO binary function.

After entering the parity bits at their bit positions, the PC parity verification bits (one bit for each code) are calculated as follows:

63 pe = XOR Word code 8it {k]

kl

Once the parity bits have been calculated and entered, the interleaver data is read from top to bottom by the columns starting with the code word O, bit OY ending with the code word 15, bit 63.

[0074] Figure 6A is a high-level block diagram of the high-speed data decoder according to an illustrative embodiment of the present invention. The high-speed data decoder implements the inverse of the high-speed data encoder compression process, and the decoder includes an optional transmission decoding process 601, a frame gain decoder 610, a quantifier

5 data samples 620, a compressor-expander 630 and a buffer 640. 8 transmission decoding process 801 includes a deinterleaver 603 and an FEC decoder 602.

[0075] The operation of the high-speed data decoder is described below with reference to Figure SA. The compressed data received is optionally entered in the deinterleaver 603, which is a 16 * 64 bit block deinterleaving process. The output of the deinterleaver 603 is introduced into the

10 FEC 602 decoder, which is an extended Hamming decoder (64, 57). The Hamming decoder can correct 1-bit errors and detect 2-bit errors per block. Figure 68 shows the Hamming and deinterleaver decoding process of an embodiment of the present invention. The data is read in the deinterleaver 603 from top to bottom starting with the code word O bit 1 and determining with the code word 15 bit 63. The syndrome is calculated as follows:

fifteen Calculation of parity bits: Pi = XOR Word code Bit [k] i = 0 .. 5

(k-1) &amp; 21 * O, where "&amp;" is a bit-by-bit ANO binary function

Syndrome = concatenation P51P41P31P21P11PO

[0076] The parity check bits (one bit for each code) are calculated as follows:

63 pe = XOR Word code 8it [k]

k = l

twenty [0077] The numerical representation of the syndrome indicates the position of the bit (if any), in which a bit error has occurred. When a bit error occurs, the bit is inverted (corrected) if the parity check bit has been set for that code. Otherwise, it is assumed that there are two (or more) bit errors in the code and the syndrome is incorrect. If the syndrome is zero, no bit error has occurred. As in the case of the encoder.

25 parity bits and parity verification bits for the 16 code words can be calculated at the same time using a 16-bit wide O-Exclusive operation.

[0078] Returning to Figure 6A, the decoded data of the FEC 602 decoder consists of the encoded quantified samples and the encoded quantified gain. The quantified gain encoded is provided to the gain decoder 610 which reads the quantized gain value from a table using the

30 quantified gain coded as the index in the table. As mentioned earlier, the coded quantified gain represents a level value of a logically separated quantizer of 64 levels.

[0079] The quantized gain value is provided to the data sample quantifier 620, where it is used to graduate the values ​​of a 32-level uniform quantizer level table. The graduated table of the quantifier decodes the quantified samples encoded in a block of linear quantified samples.

35 [0080] 8 block of quantified linear samples is converted into a PCM sample block (law A or Mu law) by the PCM 630 compression and expansion process. The PCM sample block is optionally provided to buffer 640, which Provides PCM samples as a 64 kbls signal output.

The fast speed CODEC

[0081] Figure 7A is a high level block diagram of the ultrafast speed encoder. The encoder

40 Ultrafast speed performs data compression and expansion of ultrafast speed vocal modem signals. The encoder transforms the data between the 64 kb / s PCM and an encoded compressed data stream (FEC) of 62.58 kb / s. The actual compressed data rate is 56,311 kbls, and the rest of the bit stream is used for error correction data. The ultrafast speed codec is similar to the high speed codec.

Four. Five [0082] As shown in Figure 7A, the ultrafast speed data encoder of the present invention includes an optional buffer 710, a sample format 720 preprocessor Optional. a gain calculation process 722, a delay 721, a quantifier of data samples 723, and a coding process of

optional 730 transmission. The transmission coding process 730 further includes an FEC encoder 732 and an interleaver 731.

[0083] The optional buffer 710 has a predetermined number of samples to create a sample block for the ultra-fast data compression process. The sample format preprocessor 710 eliminates the law A, or any other standard transmission format of the PCM samples and also converts the values ​​of the samples to a predetermined numerical format, such as their decimal equivalents, for convenience in processing later. The gain calculation process 722 calculates the value of the quantized gain for the sample block, and the quantizer of data samples uses the quantized gain value to create a set of quantizer levels with a predetermined separation and with the values ​​of level of quantification graduated by the value of the quantified gain. The delay shows that the quantized gain value is determined before the compression process creates the encoded quantified samples and the transmission coding process 730 is used to provide error correction coding for the transmission of the encoded quantized gain and quantified samples coded.

[0084] The operation of the ultrafast speed data compression encoder is described below. 64 kbls PCM samples (law A or Mu law) are provided to buffer 710. Buffer 710 provides PCM samples as blocks of 22.5 milliseconds of samples. At the rate of 8 thousand samples / second of the PCM, each block contains 180 samples.

[0085] Unlike the high speed codec, the ultrafast speed codec does not convert the PCM samples into linear samples. Instead, the 8-bit PCM data becomes a predetermined type of format for the representation of the sample. In the illustrative embodiment, for the Mu law, no operation is required to convert to the format, but for law A, the sample format preprocessor 720 converts the samples into a predetermined level value format before the next processing of the quantifier. As is evident to one skilled in the art, the Mu law samples could become a representation of law A, or another illustrative embodiment, both formats could become a third predetermined format.

[0086] In the ultrafast speed codec, it is desirable that the type of PCM compression be the same at both ends of transmission and reception of the link. Otherwise, without further processing, the differences between the characteristics of the Mu law and the A law can produce nonlinearity in the end-to-end characteristics of the compression coding.

[0087] The sample block received in the predetermined sample format is provided to the gain calculation process 722, which finds the sample in the block with the largest amplitude value (absolute value). The amplitude of this sample determines the quantized gain for the block. The quantized gain requires 7 bits, since the amplitude sign bit is not used.

[0088] Table 5 shows how the numbers are represented in the standards of law A and the law Mu. The absolute value of the corresponding sample is determined for those respective representations and the maximum amplitude is calculated.

TABLE 5

<dl><dt>Dec number </dt><dd>Equiv Law A HexLeyA Equiv Law u Hex Ley u </dd></dl>

<dl><dt>127 </dt><dd>255 FF 128 80 </dd></dl>

<dl><dt>112 </dt><dd>240 FO 143 8F </dd></dl>

<dl><dt>96 </dt><dd>224 EO 159 9F </dd></dl>

<dl><dt>16 </dt><dd>144 90 239 EF </dd></dl>

<dl><dt>2 </dt><dd>130 82 253 FO </dd></dl>

<dl><dt>1 </dt><dd>129 81 254 FAITH </dd></dl>

<dl><dt>OR </dt><dd>128 80 255 FF </dd></dl>

<dl><dt>-1 </dt><dd>1 01 126 7E </dd></dl>

<dl><dt>-2 </dt><dd> 2 02 125 70 </dd></dl>

<dl><dt>-16 </dt><dd>16 10 111 6F </dd></dl>

<dl><dt>-96 </dt><dd>96 60 31 1F </dd></dl>

<dl><dt>-112 </dt><dd>112 70 fifteen OF </dd></dl>

<dl><dt>-127 </dt><dd>127 7F OR 00 </dd></dl>

[0089] ~ quantified gain of the gain calculation process 722 and the complementary 2 'block are provided to the quantifier of data samples 723 after the quantized gain value has been calculated, as evidenced by the presence of delay 721 .

[0090] Data sample quantifier 723 creates a new quantifier with a set of levels of

5 quantifier from the sample block of law A or law Mu. The following explanation describes how the new quantifier is determined for a sample block. The law quantifier A divides the range of input amplitudes into 7 segments, and the law quantifier Mu divides the range of input amplitudes into 8 segments. For convenience, the following explanation describes the 7-segment law process A, although the person skilled in the art will deduce that the explanation of law A can be extended to the compression of samples of

10 LeyMu.

[0091] Each segment (except the first) has a range of amplitudes that is half of the next, and each segment (except the first) has 16 quantification level values. As a result, the size of the quantification step in each segment is twice that of the previous one. Table 6 lists the segments of the law quantifier A, along with their ranges of amplitude and step size in an illustrative embodiment.

fifteen TABLE 6

<dl><dt>Segment number </dt><dd>Input Amplitude Range Normalized Range of Amplitude Normalized step size Law Code A </dd></dl>

<dl><dt>1 </dt><dd> 0..31 0..1/64 1/2048 0..31 </dd></dl>

<dl><dt>2 </dt><dd> 32..63 1/64..1/32 1/1024 32..47 </dd></dl>

<dl><dt>3 </dt><dd> 64..127 1/32..1/16 1/512 48..63 </dd></dl>

<dl><dt>4 </dt><dd> 128..255 1/16..1/8 1/256 64.•79 </dd></dl>

<dl><dt>5 </dt><dd> 256..511 1/8..1/4 1/128 80..95 </dd></dl>

<dl><dt>6 </dt><dd> 512..1023 1/4..1/2 1/64 96..111 </dd></dl>

<dl><dt>7 </dt><dd> 1023..2047 1/2..1 1/32 112..127 </dd></dl>

[0092] Samples representing the input data signal may cover the entire dynamic range of the quantifier of law A, and the quantifier of law A becomes a new quantifier by eliminating the selected levels of the levels of the quantifier of law A. The process is illustrated below when the resulting new quantifier has a uniform value separation and all segments are used to represent a block of samples. The step size of the last segment, 1/32, is the largest step size in the quantifier, therefore, all quantizer level values ​​in the last segment are maintained. The sixth segment has a quantizer level value step size of 1/64. A step size of 1/32 in the seventh segment determines that any other level of the quantifier in the sixth segment is eliminated, resulting in a step size of 1/32. Similarly, this process is repeated for the fifth to third

25 segment. The combined first and second segments only cover a range of 1/32, and therefore, none of the quantifier levels are maintained. This translates into 31 positive levels and 31 negative levels, and a zero level is maintained to separate the first positive segment and the first negative segment, giving a uniform quantifier of 63 levels.

[0093] Next, the process calculates the maximum amplitude of a sample block and determines which segment

30 of law A contains that amplitude. For that block of data, all segments greater than that "peak segment" are ignored. The step size of the peak segment defines the size of the quantizer step. Therefore, in the uniform quantifier for the block, all levels of the quantifier in the peak segment are maintained, half of the levels in the immediately lower segment are maintained, and the level values ​​of the quantifier are assigned until or either the last segment is reached or others are no longer available

35 Quantifier level values.

[0094] Figure 9 shows the method of operation of the ultrafast speed quantifier, a 128-level quantifier, of an illustrative embodiment of the present invention.

[0095] In step 904, the method receives a compressed and expanded sample block (for example, by compression and expansion of law A or law Mu).

40 [0096] In step 906, the peak amplitude sample in the block and the corresponding segment is determined. and the peak amplitude value is the peak segment.

[0097] In step 910, all the quantizer level values ​​of the peak segment are maintained.

[0098] In step 912, unless the zero level has been reached, the 16 levels of the next segment are maintained.

[0099] In step 914, unless the zero level is reached, the 16 levels of the next segment are maintained.

[0100] In step 916, unless zero level is reached, any other level (8 level values) of the next segment is maintained. .

[0101] In step 918, unless zero level is reached, four levels of the next lowest segment are maintained.

[0102] In step 920, unless zero level is reached, 2 levels of the next lowest segment are maintained.

[0103] In step 922, unless zero level is found, 1 level of the next lowest segment is maintained.

[0104] In step 924, the zero level is maintained.

[0105] Finally, in step 926, negative levels are created using equal magnitudes as positive levels, but of opposite sign, establishing a sign value.

[0106] The peak amplitude (7 bits) and 180 coded 7-bit samples comprise the compressed output of the compression process of the ultrafast speed encoder.

[0107] Returning to Figure 7A, the encoded quantified gain and the encoded quantified samples are provided to the transmission coding process 730. The illustrative embodiment of the transmission coding process 730 induces the FEC encoder 732, which is, by example, a Hamming encoder (87, 80). The Hamming code is capable of correcting a one-bit error in the 87-bit block. The FEC encoder provides uniformly compressed and quantified data samples coded with direct error correction to interleaver 731, which is, for example, a 16 · 87 bit block interleaver. Interleaver 731 provides compressed compressed data for modulation in the RF communication channel.

[0108] Figure 7B is a block diagram of the transmission coding process of the illustrative embodiment of the ultrafast speed data encoder. A block of 87 by 16 bits is displayed. Each of the 16 rows represents a single 87-bit Hamming code word. In the encoder, the data in the interleaver block is read from left to right through the rows starting with the code word O bit 1 And determining with the code word 15, bit 86. The positions of the bits (COlumnas) 1,2,4,8, 16,32 and 64 are omitted and filled with zero. The last word / column of the interleaver block receives special treatment. It only contains the data in its first 3 rows / bit positions. The remaining rows / bit positions are filled with zero.

[0109] After filling the interleaver, Hamming encoding is performed on the 80 bits of data in each row. The Hamming parity bits are entered in the positions of bits 1, 2,4,8,16,32 and 64, as shown in the diagram. The parity bits for 6 codes can be calculated at the same time using a wide 16-bit O-8 function of the DSP. The parity bits Pi are calculated as follows and as shown in Table 7.

Pi = XOR Word code Bit [k] i = 0. • 6

(k-1) &amp; 2i ~ O, where "&amp;" is a binary AND bit by bit function.

TABLE 7

<dl><dt>Parity bit </dt><dd>XOR setting </dd></dl>

<dl><dt>PO </dt><dd> 1.3.5.7•...• 85. 87 87 </dd></dl>

<dl><dt>P1 </dt><dd> 2-3. 6-7 •...• 86 -87 87 </dd></dl>

<dl><dt>P2 </dt><dd> 4-7 •...• 84-87 </dd></dl>

<dl><dt>P3 </dt><dd>8-15 24-31. ~ 7.56H63 •• 72-79 </dd></dl>

<dl><dt>P4 </dt><dd> 16-31. 48-63 </dd></dl>

<dl><dt>P5 </dt><dd> 32-63 </dd></dl>

<dl><dt>P6 </dt><dd> 64-87 </dd></dl>

[0110) Once the parity bits have been calculated and entered. The interleaver data is read from top to bottom by the columns starting with the code word O. bit 1 And determining with the code word 15. bit 87.

[0111] Table 8 shows the interleaver block. There are 88 words numbered from O to 87. The first word

not used. but the similarity with the HSD is maintained. The first word that is not transmitted. The numbers from O to

1266 they represent 1,267 bits of the 181 words. "P" in table 8 represents parity.

TABLE 8 (continued)

<dl><dt>WordBit </dt><dd>fifteen 14 13 ... two 1 OR </dd></dl>

<dl><dt>OR </dt><dd>OR OR OR ... OR OR OR </dd></dl>

<dl><dt>1 </dt><dd>PO PO PO PO PO PO </dd></dl>

<dl><dt>2 </dt><dd>P1 Pl P1 P1 P1 Pl </dd></dl>

<dl><dt>3 </dt><dd>1188 1109 1030 160 80 OR </dd></dl>

<dl><dt>4 </dt><dd>P2 P2 P2 P2 P2 P2 </dd></dl>

<dl><dt>5 </dt><dd> 1189 1110 1031 161 81 1 </dd></dl>

<dl><dt>6 </dt><dd> 1190 1111 1032 162 82 2 </dd></dl>

<dl><dt>7 </dt><dd> 1191 1112 1033 163 83 3 </dd></dl>

<dl><dt>8 </dt><dd>P3 P3 P3 P3 P3 P3 </dd></dl>

<dl><dt>9 </dt><dd> 1192 1113 1034 164 84 4 </dd></dl>

<dl><dt>10 </dt><dd> 1193 1114 1035 165 85 5 </dd></dl>

<dl><dt>11 </dt><dd> 1194 1115 1036 166 86 6 </dd></dl>

<dl><dt>12 </dt><dd> 1195 1116 1037 167 87 7 </dd></dl>

<dl><dt>13 </dt><dd> 1196 1117 1038 168 88 8 </dd></dl>

<dl><dt>14 </dt><dd> 1197 1118 1039 169 89 9 </dd></dl>

<dl><dt>15 </dt><dd> 1198 1119 1040 170 90 10 </dd></dl>

<dl><dt>16 </dt><dd>P4 P4 P4 P4 P4 P4 </dd></dl>

<dl><dt>17 </dt><dd> 1199 1120 1041 171 91 11 </dd></dl>

<dl><dt>18 </dt><dd> 1200 1121 1042 172 92 12 </dd></dl>

<dl><dt>... </dt>&lt;dd /&gt;<dt>... </dt>&lt;dd /&gt;</dl>

<dl><dt>31 </dt><dd> 1213 1134 1055 185 105 25 </dd></dl>

<dl><dt>32 </dt><dd>P5 P5 P5 P5 P5 P5 </dd></dl>

<dl><dt>Word / Bit </dt><dd>fifteen 14 13 ... two 1 OR </dd></dl>

<dl><dt>33 </dt><dd> 1214 1135 1056 186 106 26 </dd></dl>

<dl><dt>... </dt>&lt;dd /&gt;<dt>... </dt><dd> 55 </dd></dl>

<dl><dt>62 </dt><dd> 1243 1164 1085 215 135 </dd></dl>

<dl><dt>63 </dt><dd> 1244 1165 1086 216 136 56 </dd></dl>

<dl><dt>64 </dt><dd>P6 P6 P6 P6 P6 P6 </dd></dl>

<dl><dt>65 </dt><dd> 1245 1166 1087 217 137 57 </dd></dl>

<dl><dt>... </dt>&lt;dd /&gt;<dt>... </dt>&lt;dd /&gt;</dl>

<dl><dt>86 </dt><dd> 1266 1187 1108 238 158 78 </dd></dl>

<dl><dt>87 </dt><dd>OR OR OR 239 159 79 </dd></dl>

[0112J Figure 8A is a block diagram of the ultrafast speed data decoder of the present invention. The data expansion process is the inverse of the data compression process, and the decoder includes an optional transmission decoding process 801, a gain decoder 810, a

5 820 data sample quantifier, an optional 830 sample format reprocessor and an optional 840 buffer. The optional transmission decoding process 801 includes an 803 deinterleaver and an FEC 802 decoder.

[0113] As shown in Figure SA, the compressed compressed data received is provided to the transmission decoding process 801 to eliminate the transmission encoding and correct the transmission errors.

10 transmission. The transmission decoding process 801 of the illustrative embodiment of the present invention includes the deinterleaver 803, which is a 16 * 87 bit block deinterleaver. The output of the 803 deinterleaver is provided to the FEC 802 decoder, which is a Hamming decoder (87.80). The Hamming decoder can correct 1 bit errors per block.

[0114] Figure 88 shows an embodiment of the decoding transmission process of the

fifteen ultrafast speed data decoder of an embodiment of the present invention, including Hamming deintercalation and decoding. The compressed coded data is read in the deinterleaver from top to bottom starting with the word code O, bit 1, and ending with the word code 15, bit 86. Special processing is necessary for the last column / word.

[0115J The numerical representation of the syndrome indicates the position of the bit (if it exists), at which a

twenty bit error When a bit error occurs, the bit is inverted (corrected). If the syndrome is zero, no bit error has occurred As in the case of the ultrafast speed data encoder, the parity bits for up to 16 code words can be calculated at the same time using a wide O-Exclusive operation of 16 bits

[0116] The syndrome is calculated as follows:

25 Calculation of parity bits:

Pi = XOR Word code Bit [k] i = 0 .. 6

(k-1) &amp; 21, # O, where "&amp;" is a bit-by-bit ANO binary function.

Syndrome = concatenation P61P51P41P31P21P11PO

[0117] The decoded data of the FEC 801 decoder consists of the coded quantified samples and the coded quantified gain. The coded gain is entered into the gain decoder, which provides the value of the quantized gain of the 820 data sample quantifier.

[0118J The quantifier of data samples generates a search table containing the levels of the quantifier of law A (or law Mu) corresponding to the 7-bit coded samples using the value of the quantized gain (the sample of the peak amplitude of the block). The quantifier is created using exactly the same procedure as described in the Ultrafast Speed ​​Data Encoder section, in which the search table has 256 entries, with each of the entries corresponding to one of the 128 possible

values ​​of the coded quantified samples. However, the search table is used in the opposite way. Once the search table is generated with 128 entries of the possible values ​​of the coded quantified samples, the corresponding PCM samples are found in the table by advancing the corresponding coded quantified samples (7 bit codes) to the input of the table.

5 [0119) As shown in Figure 8A, if a compression and expansion law A is desired, an optional sample format reprocessor 830 transforms the decoded block of samples into a desired sample format, such as law A . Either for a law A or a Mu law, the decoded block of samples corresponding to the reconstructed ultrafast speed data samples is provided to the output buffer 840 which provides a compressed and expanded PCM signal of 64 kb / s as a signal output