Cable television both-way system

Field of technology

The present invention relates to cable television systems, namely, a method and apparatus for transmitting data through a cable television channel sensitive to interfering noise, in which data is transmitted by a number of optional non-harmonic carrier data channels which are located in a television band channel. for retransmissions from cable subscribers to the terminal location. According to the method and device with frequency selection according to the invention, the frequencies for the reverse transmissions are set automatically on a periodic basis.

State of the art

The development of cable television systems has reached a state where the possibility of a two-way flow of information is not only desirable, but is necessary in the implementation of new services. E.g. in the implementation of the pay-per-view service, where the subscriber can freely choose what he wants to watch and expect to pay for it, at least one data channel similar to a telephone transmission channel or RF channel in the direction / opposite / from the television subscriber to the terminal is required to report service usage data. Other uses of the return path include performance meter reading, alarm services, subscriber selection and voting:, collection of statistics from subscribers about tracking and buying and home. While not every cable TV operator provides the possibility of two-way transmissions, cable TV equipment manufacturers have sought to ensure reverse transmission from the subscriber to the terminal. Virtually all of these manufacturers provide the so-called a split or bi-directional system having a full frequency spectrum for reverse transmission, which includes at least the band from 5 to 30 MHz. This band contains cable television channels T7 / 5.75 - 11.75 MHz /, T8 / 11.75 - 17.75 MHz /, T9 / 17.75 - 23.75 MHz / and T10 / 23.75 - 29, 75 MHz /. These return channels, each of which has a television signal bandwidth, can be used e.g. for video conferencing, the terminal station operator uses the lower division, middle division, or upper division system for bidirectional transmission; all three types of split transmission system typically include reverse transmission in the 5 to 30 MHz band.

An article entitled Richard Cltty and Dennis Mutzbaugh's Characteristics of Bidirectional Cable Equipment, published in papers from the National Cable Television Conference in 1984, presents the results of a study of typical cable television return devices. Five main characteristics in the 5 to 30 MHz return band were analyzed. These include white Sum and funnel, intrusion of unwanted external signals, in-phase distortion caused by faulty switchgear, impulse noise from power line interference and other influences, and not linearity of amplifiers.

White noise and Gaussian noise are terms often used to describe the characteristics of random noise. White noise describes an even distribution of noise power with respect to frequency, ie a constant spectral energy density in the considered band, here 5 to 30 MHz. The components of random noise include thermal noise proportional to temperature, shot noise generated in the active elements, and nf noise, which decreases with increasing frequency. The background noise term is used to describe the constant power level of such white noise in the considered band.

This noise is transmitted by all return amplifiers, each of which adds its own noise and contributes to the noise from all branches to the line to the terminal. This addition of noise from all branches of the distribution tree in the direction of the terminal is known as noise fusion. The constant background noise power level defines the noise level that the data carrier power level should exceed.

The invention is particularly directed to interference noise that causes peaks in the spectral density distribution in the band under consideration. Interference noise impairs efficient data transmission if a known coding method such as frequency or phase shift keying is used in a single data transmission channel. Interference noise in particular depends on the four characteristics of the feedback device mentioned earlier: intrusion, in-phase distortion, impulse noise and amplifier nonlinearity.

- 3 Intrusions are unwanted external signals penetrating the cable distribution in weak points, such as broken shielding, incorrect earthing and connection of cable sheaths and defective connectors. In these weak points, RF carrier frequencies may occur, caused by transmission in, for example, the local amplitude modulation band, the civic band, the amateur band, or the local or international shortwave band. As a result, spurious noise peaks can be observed at certain carrier frequencies in noise spectral density measurements performed on intrusion-sensitive distribution cable equipment.

In-phase distortion is the result of nonlinearities in the cable distribution caused by corrosion of the connectors, creating spike diodes. The influence of these diodes in the return distribution lies in the formation of differential products of excitation frequencies as noise peaks. power in multiples of 6 MHz, ie in the considered band 6, 12, 18, 24 and 30 MHz.

Pulse noise is defined as noise consisting of pulses of high power level and short duration. The noise of the corona and jump pulses is caused by discharges on the supply line. Temperature and humidity have a large effect on the degree of corona noise, while jump noise is the result of a distribution system failure, such as a defective or cracked insulator. The resulting impulse noise spectrum can range up to tens of MHz with a sin x / x distribution.

Amplifier nonlinearities or oscillations refer to pulsed regenerative oscillations caused by marginally stable or incorrectly terminated amplifiers. The result is a comb of frequency peaks in the return distribution, the distribution of which depends on the distance between the faulty termination and the amplifier.

From an examination of typical cable distribution Citta ad. concluded that there are holes in the gaps between the peaks of the noise spectrum plotted between 0 and 30 MHz ”. They suggested that these gaps could be used to advantage when carefully selecting return carrier frequencies to these gaps.

In the following articles, published at the 1987 National Cable Television Conference and U.S. Patent 4,586,078, Citta et al. stated that a 45 kb data signal can be transmitted alternately by coherent phase shift keying (CPSK) with 5.5 MHz or 11.0 MHz carrier frequencies or in the vicinity of channels

- 4 cable TVs T7 and T8. The switch in the subscriber's terminal alternately selects the 5.5 MHz carrier and the 11 MHz harmonic for transmission.

This method of alternating bearer message transmission continues until the data is received successfully. In other words, the alternating transmission on the two carriers continues until the terminal receives an acknowledgment signal indicating the successful receipt of the message. While these carrier frequencies are chosen to avoid peaks in the noise distribution caused by interference, there is a significant concern that such a phase shift keying data sequence will collide with noise peaks in the cable network outside of Citty's research. According to Fig. 2, taken from U.S. Serial No. 07 / 188,478, dated April 29, 1988, transmission at 5.5 MHz would be virtually impossible. Noise peaks are known to appear and disappear depending on the time of day, time, and other circumstances.

Other return path arrangements for data transmission have also been tested. These arrangements include, for example, the telephone system described by Cit tou ad. as current. In other words, the return path to the cable terminal is not created by the cable distribution network at all. The control cable is intentionally dropped due to an interference noise problem in the split system or because the system is unidirectional. Instead, the subscriber's telephone line is used for data transmission. In this case, however, there is a concern that local telephone tariffs will require the payment of line adjustment surcharges if the telephone line to the subscriber's home is used for data transmission outside the normal normal telephone service. In addition, telephone lines are only available when the subscriber is not using them, which eliminates unplanned or periodic data flows.

Another known solution for data transmission involves the use of a special data channel on a carrier frequency which avoids the difficult band of 5 to 30 MHz. This arrangement, without the use of a noise band of 5 to 30 MHz, is only possible in medium-division or high-division systems.

So-called spread spectrum data transmission is a technique that has been developed for military purposes in order to communicate securely with submerged submarines. Spread spectrum derives its name from extending a relatively low bandwidth data signal to a much wider spectrum than would normally be needed

- 5 for narrowband data signal transmission.

Recently, the security benefits provided by spread spectrum transmission have not been taken into account in favor of its ability to be used in disturbed environments. E.g. transmission systems operating over power lines, where there are high levels of impulse noise caused by power lines, have proven to be only to a limited extent. An example is the power line communication system offered by Tandy Rádio Shack. However, the Japanese company HEC, Home Electronic Group, demonstrated a home bus operating at 9600 Bd with a spread spectrum over an AC line up to a distance of 200 m of power line. The HEC system was characterized as a transition between coaxial cable /e.g. cable TV / and AC wiring, common in most homes.

U.S. Patent 4,635,274 Kaboty ad. discloses a bidirectional digital signal transmission system that uses spread spectrum transmission for retransmission in a cable television system. However, this technique is very expensive compared to returning data over the phone.

The result is that despite the development of spread spectrum data distribution and other RF. methods, in the field of cable television, the requirements for high-bandwidth data transmission from a number of subscribers to a cable television terminal using a cable television distribution network that is relatively resistant to interfering noise remain.

The concept of IPPV impulse pay-per-view is well known, but we will describe it briefly here for completeness. In essence, it is a sales method in which a paying cable subscriber can individually purchase the viewing of certain program items.

Furthermore, the purchase can be concluded only on the basis of the impulse of interactions with the terminal in the subscriber's apartment. Although it is not necessary for the items being purchased to be in progress, the system is required to allow the purchase of items that are on the agenda. The purchase must be made in such a way that it does not cause any obvious delay. immediate satisfaction.

Although there are several ways to implement this method of sale, all of these methods have common requirements. Some

- Part 6 of the system must decide whether or not to allow the sale and subsequent monitoring of the program. If tracking is enabled, the sale of that particular item must be recorded and reported to something that is generally known in the accounting system so that the program vendor will eventually get paid for the transaction.

In order to report the sale of an item, the so-called save and sell technique is used. With this method of storage and handover, the terminal assumes that if the subscriber is given the IPPV option in advance, then the sale of the item is allowed. When the subscriber takes the necessary steps to purchase the item, the terminal will allow watching the program /e.g. by decoding the video signal on a certain channel / and recording the purchase of the item. The record is typically stored in a secure, non-volatile memory because it represents a payment to the program vendor.

Obviously, the seller's billing system must from time to time receive stored sales data from all subscriber terminals in order to get paid. To achieve this, the system control computer (hereinafter referred to as the system organizer) periodically prompts the terminals to send the IPPV sales data stored in the memory. When the system organizer receives the data from the terminal, the terminal acknowledges its receipt (as Citta et al. Does) and the data is deleted from the memory to make room for further sales data. The system organizer then passes the data to the accounting system and thus the IPPV sales cycle is completed.

Although IPPV feedback data considerations are important in determining the mode of reverse RF data transmission, these IPPV feedback data considerations are not the only, but the most critical given the high throughput requirements. Other requirements, such as the use of detainees for voting of subscribers, alarm against robbers, reading meters, tip from home, energy management, etc. are in addition to the data throughput requirements of the IPPV service.

As a result, the requirements for high-throughput RF reverse transmission equipment remain in the industry to the extent that it allows for a full range of services, including IPPV.

The essence of the invention

The invention relates to a device for reverse RF data transmission for periodic and prompt updating of sales records and other information from terminals by reverse RF cable transmission. The invention mainly relates to modifications of the so-called system organizer in the terminal station for receiving data, transmitted back in the reverse data path, RF receivers with frequency switching for receiving data modulated and transmitted by a number of data channels from all terminals or modules of the system at the subscribers and the user terminal or module itself.

One of the objects of the invention is that the use of RF reverse data transmission from subscribers does not require substantial changes in the accounting system.

Furthermore, the RF data transmission from subscribers should work independently of the telephone transmission, ie they should work together, side by side. The equipment for RF reverse transmission of data from subscribers should be compatible with each terminal equipment used for direct or forward transmission. Knowledge of system equipment and terminology used can be obtained from this overview:

The system organizer is the basic control computer of the cable television system. The system organizer receives input commands from both the operator and the billing computer. It generates the relevant control transactions, which are transmitted via the control transmitter by a forward cable path to the terminals. It receives feedback from a multi-frequency receiver and a processor (called an IPPV processor) and passes the feedback to a billing computer.

Control transmitters are devices for converting standard serial data of the PS-232 interface from the system organizer to a modulated RF signal for cable transmission to IPPV terminals or modules. In my cabling system, which is available from the inventors, the control transmitter may be an addressable ATX transmitter or a control unit and a terminal encoder or a combination of both. For the purposes of the present invention, the control transmitter is primarily a pass-through device and is described for completeness.

Bidirectional amplifier. These distribution amplifiers and auxiliary amplifiers amplify and transmit a certain part of the RF spectrum in the forward direction and another part of the RF spectrum in the reverse direction. This allows two-way transmissions over a single coaxial cable. Bidirectional amplifiers are also pass-through devices and are described for completeness only.

- 8 A terminal is a device that forms an interface between a cable system and a subscriber with his television set. Among other functions, the terminals perform tuning, down-conversion and decoding of video signals from the cable on an optional basis. They accept both total and addressed control transactions / ie. transactions routed either to all or individual terminals / from the control transmitter to set up and control the services they deliver. In addition, the terminal may be equipped with an internal RF module or an interface for an additional external return data module, so that the terminal or external module may have a reliable memory for storing purchased program data or other return data, the terminal or associated module comprises a frequency switching data path transmitter according to the invention. Such a beech terminal equipped with or associated with an IPPV module will hereinafter be referred to as an STT.

An IPPV module is a module associated with a terminal if the terminal does not include an internal frequency-switched RF data return path transmitter.

The IPPV processor is primarily an RF data receiver with frequency switching for the feedback data transmitters of IPPV terminals or modules. At the same time, it detects data from modulated RF signals from up to four / or more / different return data channels. It then filters redundant messages, groups the data into packets, and passes the packets over the standard RS-232 data line to the system organizer. At least one IPPV processor is required for each cable terminal.

The overall object of the invention of the device for data transfer from subscribers is easy operation, reliable operation, high data throughput, integrity and security. In addition, the invention is designed to meet three specific objectives:

1. RF data transmission equipment must be extremely resistant to the relatively high level of discrete sources of interference typical of the return channels of cable distribution networks. The interference is due to intrusion from external RF sources into the cable network, all of which merge into the data receiver.

2. The data re-transmission method must be fast enough for the operator to receive data from all terminals, even in a large cable television system with two hundred thousand terminals to one terminal in 24 hours or less.

- 9 3. Any frequency or level adjustment for individual terminals or associated modules required in the subscriber's installation must be practically automatic.

The first two objectives correspond to two functional aspects of the invention, the frequency reverse transmission method and the reverse data access protocol according to the invention. The third goal concerns the parameters of transmission technology and especially the implementation of automatic system maintenance even in changing environmental conditions.

The invention relates in particular to this third object. Namely, the invention relates to a method and apparatus for automatically selecting a transmission frequency at an IPPV terminal or module to compensate for changes in environmental conditions. In addition to respecting the environment, relocating or reconfiguring the cable network may necessitate the selection of new transmission frequencies for use by remote terminals.

The system organizer, remote terminal or IPPV module and the IPPV frequency switching processor according to the invention automatically select the optimal frequencies for the reverse transmission. The system operator shall initially select a group of frequencies from the pooled frequencies that are available, either randomly or on the basis of previously calculated statistics corresponding to the best known frequencies. As an example, we will assume that four different frequencies were selected, although any number of frequencies could be used.

The system organizer sends a live message to all remote terminals indicating which four frequencies to use. Then all terminals use these four frequencies simultaneously to send messages back to the IPPV processor. The IPPV processor determines the number of individual successfully received messages received on each of the frequencies. After a statistically significant time interval, either the IPPV processor or the system organizer or both will exclude the frequencies with the lowest number of successfully received messages and replace them with other untested frequencies or, if all frequencies have been tested, with another frequency that had the highest specified number of received messages. . In this way, the four best frequencies of all available frequencies are constantly selected.

According to another embodiment of the invention, it is possible to generate and store as many sets of frequencies as are

- 10 at different times of the day, week, etc. using the above automatic frequency selection procedure. It is then possible to use different previously created frequency sets for each time instead of regenerating these sets. This reduces the need to re-determine the optimal frequencies.

These and other features of the invention will be apparent to those skilled in the art from the following detailed description taken in conjunction with the accompanying drawings.

«

Overview of figures in the drawings

Fig. 1 is an overall block diagram of a cable television distribution network with bidirectional distribution amplifiers and splitters allowing the connection of a subscriber's cable television terminal comprising a RF transmitter according to the invention for reverse data transmission to a terminal equipped with a frequency switching receiver according to the invention. Giant. 2 is the dependence of the interference level on the frequencies in the return band 0 to 30 MHz in one typical cable television network.

Fig. 3 is a system block diagram showing several parts of the system of Fig. 1, including a billing system, a system organizer, a frequency-switched RF feedback receiver, a terminal, and its associated RF IPPV feedback module. Fig. 4 is a block diagram of a typical STT terminal that includes a command receiver addressed out of band. FIG. 5 is a block diagram of an IPPV module for the terminal of FIG. 4; this module contains either a part of the terminal or is connected to the terminal by the appropriate bus system. Fig. 6 is a circuit diagram of the BPSK modulator of the IPPV module of Fig. 5. 7 is a timing diagram of the sequence of feedback data from the frequency-switched RF reverse data transmitter of FIG. 5. FIG. 8 is a block diagram of the processor / receiver / IPPV indicated in the system diagram of FIG. 3. FIG.

to 13 are block diagrams of several IPPV processor subassemblies according to FIG. 8. FIG. 9 shows the input module, Qbr. 10 is a frequency receiver, FIGS. 11A to C are an RF receiver, FIG. 12 is a signal strength analyzer, and FIG. 13 is a control unit assembly. FIG.

- 11 Fig. 15 is a timing diagram of an IPPV data transmission sequence. Fig. 16 is a timeline demonstrating the basics of Miller coding. Fig. 17 is a Karnaugh map showing a suitable function to be performed according to the state of the reverse data channel.

Examples of embodiments of the invention

Giant. 1 shows a typical cable television distribution network 100 for distributing cable television signals to subscribers and for receiving feedback messages from the subscriber terminals 120. The cable television distribution network 100 connects the terminal 110 to a number of televisions 130 via cable television terminals 120. The cable television network 100 is connected in a tree configuration with branches 148 and 150 using splitters 143. Occasionally, bypass switches are used in place of splitters to switch transmission between the end station and subscribers to only one branch of the input to the splitter 143. One of the objects of the invention is to obviate the need for bypass switches previously used to increase data throughput from subscriber to end station. In the forward direction, many subscribers typically receive the same signal transmitted from the IIP terminal, typically a broadband cable television signal. In future systems with increased bandwidth, such as fiber-optic systems, it is unlikely that different subscribers would be able to receive different signals for them only, a scope previously reserved for telephone companies only. The distribution amplifiers 142 are also regularly distributed in the cable network 100 to amplify and repeat the transmitted signal. The transmission from the subscriber terminal 110 to the cable terminal 120 is sensitive to interference entering the main line 141, branches 148, 147, 146, 145 and the connection 144. However, far more severe interference enters the transmission from the subscriber to the terminal 110.

The frequency-switched reverse RF data transmitter 200 may be within or associated with the cable terminal 120 and allows communication with the terminal 110 by transmitting messages back over the cable television network. The terminal 110 includes a frequency-switched RF receiver 300 for receiving messages transmitted by the RF reverse data transmitter 200 at the cable television terminal 120 or the IPPV module located at any location of the entire number of subscribers. Other subscribers equipped with IPPV or other services requiring data return may be equipped with telephone transmitters for transmission to a telephone processor (not indicated) at the terminal.

A set of cable television networks are so-called split systems equipped for two-way transmission, ie transmission from the terminal to the subscriber and from the subscriber to the terminal. In these cable television networks, amplifiers 142 are equipped for bidirectional transmission, including reverse amplification. Cable television companies have so far avoided the use of two-way transmission in a cable television network, partly because the return transmission from the subscriber to the end station is significantly more sensitive to interfering noise. The feedback is more sensitive to interfering noise because the cable network has the shape of a tree configuration, allowing the interference to propagate and amplify from all points of the cable network in the reverse direction. This can be referred to as a funnel phenomenon. E.g. disturbing noise 160^and 1-61 ^on at line 144 and branch 154 will combine to noise 162 in splitter 143 connected to line 144 and branch 154. As the signals propagate toward the IIP, the noise will combine with noise at branches 153, 152, 151. 150 and on all other lines in the entire cable television network. In the reverse direction, it may become difficult to distinguish the transmitted data signal at the IIP terminal from the noise induced in the branches of the cable television network.

Interfering noise can include impulse noise, in-phase distortion of external signals, and amplifier nonlinearity. Examples of sources of interfering noise are lightning IP, radio 11 and distribution network 12. Cable television networks may contain old and poorly grounded and connected cable shields, etc., which allows noise to enter anywhere in the cable television network. Aging distributors 143 or old non-linear distribution amplifiers 142 can also cause interfering noise. Because the interfering noise from each individual cable extension branch affects the reverse transmission, while the direct transmission is only affected by the interfering noise along a single line /e.g. branches 141, 148, 147, 146, 145 and connections 144 / the cable television return network requires

- 13 aging costly maintenance more often than the forward network. The invention enables the transmission of feedback signals by an imperfect cable television network where reverse transmission has hitherto been difficult without costly regular preventive maintenance of the cable television network. The invention allows the two-way transmission of messages over cable television networks even at a higher level of interference than has hitherto been possible.

Fig. 2 is a graphical representation of the noise energy level versus frequency in a typical cable television network. The measurements were performed during the main monitoring period / evening / on a relatively new installation. The effect of interference is probably the most serious from a local station with amplitude modulation at 1500 kHz, the BBC World Service, the Voice of America and amateur broadcasting at 21 MHz. It is immediately apparent that transmission by known techniques in the T7 channel (5.75 - 11.75) would be practically impossible. Furthermore, it can be seen from the layout that the higher the frequency, the less difficult the interfering noise ·

At the time of measurement, the effect of in-phase distortion was not very significant. However, the network was examined again a year later and peaks multiplied by in-phase distortion appeared at 6, 12, 18 and 24 MU », as expected.

Fig. 3 is an overall solution of the IPPV system according to the invention. The system includes an accounting computer 305 that records and maintains records for each subscriber in the system. Records typically contain information such as the subscriber's name, address and telephone number, the type of equipment the subscriber has and which paid services the subscriber is authorized to monitor. A cable operator typically either owns a billing computer, leases equipment from a company that specializes in this type of equipment, or shares time on a computer that belongs to the billing company.

The billing computer 305 is connected to the system organizer 110. The system organizer 310 controls the operation of the cable system. The system organizer 310 maintains a list of all addressable terminals in the cable system as well as the services that each terminal is authorized to receive. The system organizer 310 also determines and maintains the parameters selected by the operator for each system. These parameters may include the frequencies associated with each cable channel in the system, which channels are being coded, the security measures of the system and the system time. In addition, the system organizer 310 is responsible for authorizing and disallowing pay-per-view programs.

The system organizer 310 also stores IPFT information. The system organizer resident program reads IPPV transactions selected from cable system terminals. Transactions are stored in the system organizer's database until they are retrieved by accounting computer 305. System organizer 310 controls the transmission of IPPV purchase feedback information by sending data requests to cable system terminals.

As indicated in Figure 3, the commands generated by the system organizer can be transmitted to the terminals in one of two ways. In the first method, the addressable ATX transmitter 314 transmits commands from the system organizer 310 (or via the terminal control unit 312) on a dedicated channel (e.g. 104.2 MHz / in a format understandable to addressable terminals. In the second method, the commands are transmitted using a so-called in-band system, where the commands are incorporated into the video signal by the activities of the encoder 313. A system operating in the band? is described in the co-filed dependent application no. No. 188,481, incorporated herein by reference. Other methods can be used for both addressed or global data transmission from the end station to the subscriber's terminal, and the present invention is not to be construed as limited in this regard. E.g. acoustic transmission, spread spectrum or other means of transmission over the same cable may be used, or an equivalent set of alternatives may be used on a switched or private telephone line or on a distribution line.

The subscribers of the cable system can be equipped with terminals 315 FIG. 3 shows three terminals, two of which / 315a. 315b / are associated with an in-band system and one (315c) is associated with an out-of-band system. E.g. terminals 315a and 315b may be Scientific Atlanta Model 8570 and 8590, while terminals 315c may include Scientific Atlanta Model 8580. The terminal allows the subscriber to tune and decode services required by the cable system operator. Each terminal contains a single digital identifier as a numeric address that allows the operator to send commands directly to each terminal. These commands are called addressed commands. Terminals are also able to receive

- 15 global commands, processed by all terminals in the cable system. Subscribers who are entitled to purchase pay-per-view pulses are equipped with terminals with built-in pulse modules. In summary, the impulse module allows the subscriber to authorize his terminal to receive the program on the basis of payment for monitoring, storage of data associated with the purchase of the meeting and transmission of stored data to the cable network operator. As indicated in FIG. 3, the stored data may be transmitted back to the operator by the telephone pulse module using the public switched telephone network 317 via the telephone processor 321 or via the RF pulse module using the RF callback 319 via the IPPV processor 322. The RF callback will be described in more detail below. The telephone processor 321 and the IPW processor 322 are connected to the system organizer 310 by a suitable interface such as an RS-232 interface.

The accounting computer 305 sends a transaction to the system organizer 310, which identifies whether the terminal in the system uses RF call 319 or the telephone network 317. The system organizer 310 then enters the transaction into the terminal 15. to run and set it. E.g. the frequencies to be used for RF transmission and the calibration procedures described below must be entered into the RF pulse module. These frequencies may be entered into the module during production or may be entered there by a global transaction from the system organizer 310. Alternatively, these frequencies may be entered by an addressed command.

Giant. 4 shows a block diagram of a conventional addressable terminal of the known solution, namely the Scientifio Atlanta 8580 terminal. According to the principles of one embodiment of the invention, the terminal is a through-going device and does not form part of the invention. With its output, the first microprocessor 400 only transmits all commands received by the addressable data receiver 430 to the second microprocessor 504 in the IPPV feedback module indicated in FIG. 5 via the connector 490. In an alternative embodiment, the functions of the second microprocessor 504 of the module of FIG. 5 may be integrated into the first microprocessor 400, in which case a microprocessor with a larger capacity than M50751 will be required.

The basic building blocks of the out-of-band addressable terminal are the converter and tuner 410 for receiving and converting the incoming signal from the cable. The data receiver 430 receives the converted 104.2 MHz microband signal or other suitable carrier frequency from the converter and tuner 410. The output of the converted television signal is decoded by the decoder 420 as needed. other subscriber equipment / not indicated /.

Associated with the microprocessor is a first non-volatile memory 470, timing logic 480, direct input keypad 440, infrared or other remote control receiver 450 for receiving remote control input, and a display 460. The display shows, for example, the tuned channel number or time of day.

According to the invention, the described terminal model 8580 is only a through-going device. Each of the 8570, 8590 and other third-party terminals normally includes a control processor such as the first microprocessor 400, which must always have inputs and outputs or connectors for exchanging data with the module as shown in Fig. 5 or for controlling the elements of Fig. 5. if this module does not contain a microprocessor. The second memory 502 of FIG. 5 is a nonvolatile memory that simply replenishes the capacity of the first memory 470. The first microprocessor 400 has access to it.

For the implementation of domestic purchasing, energy management, meter reading, burglary alarm and other services in addition to the IPPV service, the terminal must contain appropriate interfaces for entry and exit to various devices in the subscriber's household / none of which is indicated in Fig. 4 /.

Fig. 5 shows a group diagram of an IPPV module according to the invention. The IPPV module is a BPSK microprocessor transmitter used to transmit information in the reverse direction of cable television networks from the subscriber's site to the terminal. The second microprocessor 504 communicates with the first microprocessor 400 of the terminal and receives information that is stored in the third memory 503 (for later transmission) or transmission instructions. During the transmit cycle, the second microprocessor 504 turns on the power to the frequency synthesis circuits, programs the appropriate transmit frequency, turns on the power amplifier, sets a predetermined gain level at the modulator, and transmits the required information.

The second microprocessor 504 is the brain of the module because it determines when

- 17 is to be transmitted (according to the instructions transmitted from the terminal station, which we will discuss below), determines and sets the frequency and power level of the transmission and encodes the data stored in the third memory 503 for transmission. In order to ensure a prompt and efficient return of the data, this data is preferably pre-formatted when stored in the third memory 503.

Upon completion of the transfer, the second microprocessor 504 also turns off the RF circuits, thereby reducing the noise output of the module and reducing the overall power consumption. Third memory 50? stores consultation data (pre-formatted for transmission), security information, transmission frequencies and power levels, and module identification information. Memory £ 50? it also stores program viewing statistics, as will be described in more detail below.

The phase-locked loop 505, low pass filter 506, and voltage controlled oscillator 507 synthesize the frequency to be used for transmission. The frequency is synthesized from a 4 MHz crystal clock 501, which also controls the second microprocessor 504. This arrangement reduces the number of components required to perform the synthesis and also eliminates the problems that could arise when using a dual clock of the same frequency.

The module phase lock loop 505 receives serial data from the second microprocessor 504 to set its registers to a certain frequency. The phase-locked loop 505 compares the sampled signal from the voltage controlled oscillator 507 with the signal derived from the 4th MHz clock 501 to determine if the generated frequency is higher or lower than the programmer frequency with a polarity representing a high or low generated frequency. The low pass filter 506 performs mathematical integration of this signal and generates a voltage to control the output frequency of the voltage controlled oscillator 507. The output of the oscillator 507 is fed to the modulator 508 and also back to the phase locked loop 505 so that it can be sampled again and the process repeated. broadcasting.

The data filter 510 is a bandpass filter that prevents the RF energy of the transmitted digital information from being modulated on the RF carrier. Thus, the data filter 510 acts as a modulation energy limiter of the modulated signal within certain limits.

The modulator 509 receives the input of the filtered data from the second microprocessor 504 and the RF carrier from the voltage controlled oscillator 507 and modulates the phase of the RF carrier proportional to the data signals. The modulator also uses a DC bias, generated by a digital-to-analog resistance converter, to control the overall gain of the modulated signal. The converter is controlled directly by the second microprocessor 504. The modulator 508 is described in more detail below in connection with FIG. 6.

Three modulation solutions have been considered in the present invention for RF feedback data: Binary frequency shift keying (FSK), binary phase shift keying (BPSK) and direct spread spectrum (DSSS) with BPSK modulation. * Many solutions were considered too complex and unnecessary. as bandwidth saving is not a critical requirement.

Of these three options, HPSK has the greatest resistance to broadband noise, DSSS has the greatest resistance to discrete frequency interference, and FSK is the simplest to implement. On the other hand, BPSK and FSK malm have resistance to signal interference at the same frequency, but the receiver for DSSS is quite complex and has a wide noise band. The DSSS transmitter also requires a very complex filter to prevent interference with both the forward and reverse video signal. The FSK receiver is suffering (in this case) from the capture phenomenon, which is a problem in this situation.

The system of the invention provides some of the best features of all. The system uses HPSK signaling on four different frequencies. This solution can be called HPSK with frequency selection / or FD BPSK /. In this way, the noise band of the receiver is very narrow, the inherent properties of BPSK are used in noise suppression and, with the appropriate choice of frequencies, interference by discrete frequencies is avoided. However, although BPSK modulation has been used in the present invention for the above reasons, other modulation methods may be used, and the invention is not limited thereto.

The power amplifier 509 amplifies the resulting signal from the modulator 508 to the desired output level of the module. The gain of the amplifier is at a fixed level, with the signal from the crosstalk control 513 controlling the on and off of the power amplifier 509.

The crosstalk control 513 is a circuit allowing the second microprocessor 504 to control the state of the power amplifier £ 09. In the event of a failure of the second microprocessor 504, the anti-crosstalk control 513 turns off the power amplifier 509 after a predetermined time or after several consecutive transmissions. This prevents the module from transmitting

- 19 messages longer than suggested or more frequent than intended regardless of the state of the microprocessor. Terminals that chatter or scream are unmanaged and generate intrusive messages that can block the entire system when enabled. The stereo circuitry prevents blabbering by turning off the data transmitter after a predetermined time that is longer than the longest message would need. Anti-custody control is described in commonly assigned U.S. Patent No. 4,692,919, which is incorporated herein by reference.

The pooling filter 511 is a two-part filter: a bandpass filter 515 for 12-19 MHz to suppress the harmonic energy of the module transmitter, and a high pass filter 516 54-870 ItiRz for cable television signals to pass to the terminal undisturbed.

The design considerations associated with the solution of the IPPV module for the so-called internal systems are not particularly suitable for the solution of the so-called external systems. For example, internal systems refer to in-band and out-of-band addressable terminals such as Scientific Atlanta 8570, 8580, and 8590 terminals. External conditions involve the relocation of a terminal device from a subscriber's premises. Such external systems include, for example, the prohibition and restraint technique. As a result, there is e.g. at least a domestic, if not connection, cable separation between the cable terminal and the subscriber's equipment, which is not particularly suitable for data transmission. On the other hand, some external subscriber equipment is needed for IPPV, home shopping and two-way services that cannot be implemented with a conventional television set.

As a result, it would be difficult to implement the module of Fig. 5, which pre-documents a bus or other transmission path between the terminal and the module, with conventional house cables or connections without any special transmission solution. The invention therefore relates to those principles of the terminal and module solution which can be extended from the internal terminal solution to the IPPV module solution for subscriber units in the so-called external prohibition and hold system.

Fig. 6 is a detail of the BPSK modulator of Fig. 5. BPSK modulation is a modulation method that alternates the phase of the RF carrier frequency between two possible states to display one of the two logic states. The BPSK modulation method used in the IPPV transmitter of the present invention involves the use of a balanced differential amplifier to generate phase changes in the RF carrier to display the encoded digital information. Although there are a number of possible solutions for implementing this type of modulator, the use of a differential amplifier, as indicated in Figure 6, also provides a means of varying the overall gain of the circuit and thus allows microprocessor control of the output power level. By supplying a constant level RF carrier based on transistor Q3 in FIG. 6 and by combining this signal with a DC bias from a digital-to-analog converter controlled by a second microprocessor 504, a pseudo-linear output power control is built into the inexpensive BPSK modulator.

The BPSK 600 modulator includes a resistive network 602 of programmable gain control. This gain control resistor network 602 comprises four resistors R1 - R4 1 kQ, 2.2 kQ, 3.9 kQ and 8.2 kQ. One end of each resistor R1 to R4 is connected individually to inputs B3 to BO. The other end of all resistors is connected to a common output 605. The output 605 of the gain control resistor network 602 is connected to the base of transistor Q3 via a resistor R5 of 3.3 kQ. A voltage of 5 V is applied to the first point between the output of the Gain Control Resistor Network 602 and the resistor R5 through a resistor R6 of 3.3 kQ. Oscillator output 50? /giant. 5 / is connected to the base of transistor Q3 via capacitor C2 10 nP.

The emitter of transistor Q3 is connected to ground via resistor R7 8.2 kQ. The point measi by the emitter of transistor Q3 and resistor R7 is connected to ground through capacitor 03 10 nP and resistor R8 33 Q.

The emitter of transistor Q1 is connected to the emitter of transistor Q2. The collector of transistor Q3 is connected to the connection point of both emitters. The input data is fed on the basis of the transistor Q1 through the data filter 510 / FIG. 5 /. The point between the data filter 510 and the base of transistor Q1 is connected to ground via capacitor C4 10 nF and via resistor R9 27 kQ with resistor R10 27 kQ. Line A shows the connection between the two points.

The point between resistors R9 and R10 is connected to ground via resistor RU 12 kQ and via resistor R 12 3.3 kQ with input + 9 V. The point between resistor R 10 and the base of transistor Q2 is connected to ground via capacitor C5 10 nF,

The collectors of transistors Q1 and Q2 are individually connected to the primary terminals of transformer 650. The center of the primary winding of transformer 650 is connected to + 9 V via resistor R13 47 Q. One terminal of transformer 650's secondary winding is output of modulator and the other is via capacitor 06 10 nF connected to ground.

We will now explain the function of the modulator 600.

The modulator 600 receives data input from the second microprocessor 504 of FIG. 5 and filters this data to limit the high frequency content. The filtered waveform changes the collector current of transistor Q1 to one of two possible states, displaying either the number one or zero. The base of transistor Q2 is kept at a constant voltage.

The RF oscillator forms the base input of transistor · Q3. The collector current Q3 is maintained at a constant level determined by the voltage output of the digital-to-analog converter in the form of a resistive network 602 of programmed gain control. Because the collector current Q3 is kept constant, the total emitter current of transistors Q1 and Q2 i

must be equal to the current of transistor Q3. The collector current Q1 changes in proportion to the data signal based on it and thus the collector current Q2 changes in the opposite way so that the total current is constant. The RF current from collectors Q1 and Q2 forms a differential voltage across the primary winding of transformer 650. The differential RF signal with transformer 650 is converted to a simple signal, producing an RF carrier frequency that changes polarity / phase inversion / proportional to the Q1-based data signal. This is a BPSK signal that is amplified and transmitted.

The gain control function of the modulator is the result of the bias voltage applied at the base of transistor Q3. This DC bias in combination with the RF signal from the oscillator determines the collector current (and gain level) proportional to the bias voltage. Thus, as the bias level increases due to the output of the programmable control resistor network 602, the gain of the RF signal by the transistor Q3 also increases. The programmable control resistor network 602 is designed to have a complementary dc response to the digital input and to cause a linear increase in the RF power at the modulator output. In other words, for each increment of increase of the four-bit digital signal, the output power of the modulator increases by a fixed value.

We will now describe the operation of the individual components previously described in connection with the features of the invention.

As mentioned earlier, so that each terminal 315 can transmit

- 22 purchase purchase information back to system organizer 310, must have a reverse path / as opposed to forward path used to send control information from system organizer 310 to terminal 315 / »As mentioned, the use of an IPPV RF system in cable networks that have the ability to divide return channels. These cable systems have amplifiers that allow the propagation of channels T2, T8, T9 and T10 (approximately 0-30 MHz) in the reverse direction, i.e. to the terminal.

The invention proposes an IPPV module according to Fig. 5, which uses parts of the T8 channel for transmissions from IPPV terminals or modules to a frequency-switched data receiver at an end station via an optional series of modulated data carrier channels. The use of T7, T9 and T10 signals for video conferencing or other transmissions is not adversely affected by data transmissions, which are generally limited to the T8 channel band.

The use of return channels in the cable network as a data transmission network for subscriber terminal information suffers from two fundamental shortcomings: the high noise and disruptive return environment, as discussed in detail earlier, and the lack of a concurrency mechanism by which data can compete for network access. Both problems arise from the topology of the system, which is an inverted tree, as indicated in Figure 1.

In terms of interference, tree branches can act as a large antenna network. Faulty shielding θ broken or loose connections of the cable system allow the penetration of RF interference into the system, as we described earlier. Because the amplifiers are set to total unit gains, in-band interference and noise are regenerated in each of the amplifiers. In addition, in the reverse direction, the interference and noise from the individual branches in each network node combine. As a result, all the interference and noise that has penetrated the entire cable system is finally added up at the end station where the IPPV RF data receiver is located. To minimize these problems associated with the use of cable TV return channels for data transmission, a number of four channels in the range of 23 data channels of 100 kHz in the T8 TV channel band was chosen for use in this IPPV RF system, based primarily on data throughput considerations. As described below, the invention is not to be construed as limited to four channels, but may be used

- 23 and more than four channels. The probability of receiving messages increases with each additional channel used, but the cost of additional transmitters and receivers for additional channels becomes prohibitive.

The return channel vi deo 6 Kz can be divided into 50 transmission channels of 100 kHz each, of which 23 are used in the current implementation. Four of the twenty-three channels are selected according to the frequency of noise and interference. Both transmitters and receivers are frequency variable. The frequencies used for reverse transmissions can be automatically programmed by the system organizer computer to avoid channels that contain noise or major interference. These frequencies can be changed as often as necessary to adapt to time-varying interference.

Each transmitter will sequentially transmit its data at a rate of preferably 20 kbps on each of the four frequencies. Four RF receivers (one tuned to each channel) are used in the terminal. This arrangement provides redundancy for each message. The probability of an error caused by interference on the same frequency is now the product of four probabilities that interference is present on all four channels at the time this channel is used by the transmitter.

Note that this may provide better parameters than spread spectrum systems, because the step-by-step arrangement provides both time and frequency variation.

In a typical feedback system, four video channels are available: T7, T8, T9 and T10. Usually, the lowest T7 channel has the most noise and the highest T10 channel has the least noise. This would lead to the conclusion that the T10 channel would be the best. However, there are other considerations.

Many cable operators use or are required to keep several return channels on standby. These are sometimes used by her for video conferencing, shared access television, character generator connections to end stations and modem service. Because video is much less tolerant of noise than data transmission, it is desirable to keep the cleanest channels free and use one of the lower channels.

Data obtained from direct observation of the feedback networks of several customers show a significant deterioration of the channel quality between T8 and T7.

Although the BPSK system could probably operate on channel T7, it will generally be much easier to place pure frequency bands in channel T8.

- 24 The last factor, concerning the choice of frequency, is the location of the harmonic frequencies of the transmitter. It is important to keep the second and third harmonic transmitters out of both the upper return channels and the forward video channels. If the transmitter frequencies are limited to the band 14 to 18 MHz, the second harmonic / 2xf ° / will be between 28 and 36 MHz and the third harmonic / 3xf ° / between 42 and 54 MHz. The second and third harmonics will therefore be outside both the forward and reverse video channels / above T10 and below channel 2 /. This significantly reduces the requirements for filtering the transmitter output and at the same time the costs and increases the reliability. Thus, by selecting the T8 channel, unlike dtty ad. we intentionally avoid carrier frequency harmonics, which could negatively affect the reverse transmission if odd and even harmonics fell in the upper part of the 0 to 30 MHz transmission band.

The source of penetrating interference is typically discrete frequencies, which are time-varying in nature. Although averaged measurements by a spectrum analyzer may indicate areas or bands in the T8 channel that may be completely undesirable at some point in time, it is still difficult to predict with certainty which frequency or frequencies can always be used. However, at any given time, there is typically a significant band within the T8 channel with a fairly low level of noise and interference, allowing reliable transmission. The proposed RF IP frequency switching system is designed to take advantage of this fact in preventing interference in several additional ways: minimum bandwidth data transmission method, frequency switching, several / current / transmission channels and time-randomized redundant message transmissions.

The RF module of Fig. 5 transmits IPPV event data on up to four different channels (frequencies) each time it attempts to return data (or repeats it). The actual number of frequencies used is programmable according to the terminal in a given implementation from one to four, although the invention is not limited in this regard. The frequency-flexible nature of the system allows the feedback system to be programmed to operate on channels / frequencies / that do not have permanent strong interference. In addition, the use of several frequencies eliminates random and time-varying sources of interference.

E.g. after the initial assembly of the system, it is possible to use a spectrum analyzer and find several usable 100 kHz channels

- 25 in the frequency band 15.45 - 17.75 MHz, which have an average low level of interference. However, at any given time, there is always a certain probability that a random or time-varying source of noise can interfere with the feedback. The probability of interference in one channel is relatively independent of the interference occurring in another (not adjacent) channel.

To illustrate, assume that the probability of harmful interference occurring during a transmission on a channel is 50%. · Thus, it is not possible to use more than one-half of the bandwidth of any channel. From another point of view, the probability of passing a message with back data is only 50%. However, if we attempt to send a message through a number of channels at the same time, the attempt will only fail if all attempts on all channels are unsuccessful. In other words, the only case where none of the attempts to transmit the message will be successful is when all four attempts are unsuccessful. The probability of such a case when using four channels will be:

0.5 x 0.5 x 0.5 x 0.5 = 0.0625 / 6.3% / or only one-eighth of the probability of 50% failure using one channel. In general, if the probability of failure is due to interference on one channel K, then the probability of failure when using four channels is K⁴. The relative improvement is therefore K / K⁴ chili l / K³.

The system organizer, the IPPV processor and the terminal module store two sets of / to / four usable channels in the designed embodiment. We call these two sets of channels Category 1 frequencies and frequencies

2. category. It will be apparent to one skilled in the art that the invention is not limited to two categories of frequencies, where each category contains four frequencies. In fact, any number of categories can be used, and each category can contain the same or different numbers of frequencies. Commands from the system organizer to the IPPV processor and the terminal can immediately switch functions from one of the seven operating frequencies to another. Another option is to program the system organizer to automatically cycle the system between categories at different times of the day.

In the proposed arrangement, two operating modes are still immediately available without interruption. E.g. Category 1 can specify three channels for return data and one channel for automatic calibration of the terminal module, while Category 2 can define

- 26 four channels for data transmission. During the day of installation, the system can be programmed to use Category 1 so that automatic calibration can take place. During the night, the system can be programmed to use Category 2 to maximize the benefits of multiple return data channels.

If it is known that the quality of certain return channels changes significantly during a certain time of day, two categories can be used to quickly and automatically switch several channels at pre-programmed times. E.g. due to the interfering radio transmitter, channel A may be much better than channel B at 04.00 to 18.00, but somewhat worse than channel B at night / from 18.00 to 04.00 /. It is then advantageous to assign channel A to one category and channel B to another and program the system to switch at 04.00 and 18.00 to the appropriate category.

Assuming low noise in a number of channels, a lower number of return data channels can be used without reducing data throughput. In this way, different groups can broadcast on different channels within one category.

The IPPV processor, together with the system organizer, collects and maintains statistics on the number of valid different messages received on each of the four RF channels. The number of messages sent by the terminal on each channel used is essentially the same. Thus, after a statistically significant period of time, the number of valid messages on all channels used will be the same if the quality of all channels is equivalent. Conversely, if the quality of one or more channels is poorer, the number of valid messages received on these lower quality channels will be either lower than the number of messages received on so-called cleaner channels.

It follows that the cumulative sums of the various messages received for each channel are an excellent indicator of the relative quality of the channels. Quality can be compared between channels in the short term as well as · analyze long-term trends on individual channels.

Although the proposed layout only allows the display of cumulative sums of messages during each callback zone, this information, together with other system features, can be used to automatically select a frequency. E.g. the following algorithm would eventually test all channel frequencies and use the best four:

1. Start with four seemingly "good" frequencies

- 27 2. Analyze data feedback performance for a statistically significant period of time

3. Remember the relative quality of the worst frequency and take it out of use

4. Mahraa the worst frequency at a frequency not yet tested

5. Repeat steps 2 through 4 until the order of all available frequencies is determined

6. Continue to use this algorithm, except that you select the best rated frequencies if a substitution needs to be made.

This algorithm can be easily adapted to use more or less than four channels.

The IPPV system of the invention uses Bffiller data coding with carrier modulation by binary phase shift keying (BPSK). Bffiller data encoding gives excellent information about the timing of regenerated data using the minimum bandwidth.

When the terminal receives a request to send back data from the system organizer, the message tells the terminal which frequency category to use, how many times / N / to send the message, and how long the transmission period is. The terminal then calculates ΓΓ pseudo-random times to start the message in the specified transmission period for each of the frequencies used. The feedback data is then sent up to N times on all frequencies. The start time is calculated independently for each e-frequency, so both the message start time and the frequency order are random. The transmission of each message at a random time on a certain frequency is primarily a function of the method of statistical access to the medium used (see the next section on the non-medium access protocol). Message redundancy provided by multiple transmission attempts at multiple transmission frequencies is the primary factor enabling resistance to intruding noise. This technique is essentially a spread spectrum and frequency hopping system, although the hopping is slow relative to the data compared to the known spread spectrum.

To take advantage of the ability of terminal transmitters to transmit on multiple frequencies, the IPPV processor includes four separate receiving sections that can receive messages simultaneously. The system organizer sets a frequency category at the beginning of each return data arrival period to ensure that they match the terminal frequencies. The microprocessor control unit of the IPPV processor decodes the messages from all receivers. The messages are organized into packets and passed to the system organizer. The IPPV processor controller also sorts the messages to exclude redundant messages received from the terminals during each transmission period.

In the operation of an IPPV system, it is generally desirable to be able to request a message with feedback data, ie to select terminals equipped with IPPV modules according to several different criteria. The following list summarizes the most useful cases for requesting feedback from certain groups of terminals:

1. Unconditionally, ie all terminals need to report.

2. All terminals containing IPPV data for one or more programs.

3. All terminals containing IPPV data for a specific program.

4. Certain terminals individually, regardless of program data.

Furthermore, as mentioned earlier, it is very important, even in the first case of unconditional data request, that all terminals be able to return data in a period of less than 24 hours. This should be possible in the case of a number of terminals a thousand or even several hundred thousand, which leads to a throughput requirement of around 25 thousand. IPPV data response per hour.

Each ae of the reverse narrowband data channels can transmit only one message at a time. This means that when two or more terminals anywhere in a given cable system send messages that overlap in time, the transmissions will interfere and all data affected by the collision is likely to be lost. Therefore, in three of these cases, some medium access control procedure is needed to prevent several terminals from trying to use the reverse data channel at the same time.

However, all these cases could be dealt with as a series of individual data requests / as in the fourth case /. However, this is not compatible with throughput request due to the systemic delay of the messages associated with a typical request-response message sequence. It is much more efficient to send a single group request for data to a relatively large group of terminals, which will then return the data oodle scheduled

- 29 procedure or media access protocol. This protocol must ensure a high success rate of messages, ie without collisions.

Unfortunately, known media access protocols, such as those used in local area networks and which rely on bearer sensing mechanisms to prevent collisions, are unsuitable for use in a cable system. The topology of the inverted cable system tree adds the transmitted signals from the individual branches and propagates them towards the terminal. Terminals located in different branches, each of which is isolated by a node amplifier or other device, cannot detect the presence of an actively transmitting terminal in another branch.

Another access protocol, time interleaving, also suffers from time / fluctuation delay / worst case messages. This forces the time slot for each terminal to be unbearably long, which reduces throughput.

All these circumstances have led to the development of a medium access protocol that provides acceptably high throughput at the calculated collision tolerance. The method uses the predicted statistical probability of collisions (and, conversely, the successful passage of messages) for a given controlled, evenly distributed frequency of terminal attempts to transmit back data.

To be very simple, according to this method, the system organizer sends data requests to manageable large subgroups of the total number of terminals. / These subgroups are independent of the four election cases mentioned earlier./ Each subgroup or simply group has a defined time period in which to return data. Within this period, each terminal independently selects a programmed number of pseudo-random moments to start the reverse data transmission. When using relatively large subgroups, return attempts are statistically evenly distributed over a period of time. Furthermore, because the mean frequency of attempts is predetermined and the mean length of the feedback messages is known, the resulting probability for at least one successful feedback data message for any terminal can be predicted.

Γ While this statistical concept is the basis of the data return method, a number of other key elements are needed to make the process operational. We can summarize them as follows:

- 30 1. Determine the optimal frequency of trials that gives the best effective throughput of feedback.

2. The total number of terminals at the cable system terminal is divided into manageable subgroups of known range. The size and number of subgroups as well as the data return period can be determined for a given optimal number of trials.

3. A data return plan is needed that provides a structure for how the system organizer will request feedback from each group.

4. The set of rules controls how the terminals in the groups respond to data return requests and dst acknowledgments within the return data sequence.

Giant. 7 shows the time course of a typical sequence of feedback data. As mentioned, the total number of terminals is divided into manageable subgroups of approximately the same size. We will simply call them groups. The time for which each group is allowed to transmit feedback data is called the group period (or short period). During the IPPV data lookup, the system organizer sequentially sends a data request to each of the groups at the cable system terminal. One complete sequence of data returns from all groups is called a cycle. Finally, a sequence of two or more cycles that produce a complete (typically daily) data return sequence is called a zone. If a terminal returns its data during a given zone and receives an acknowledgment, that terminal will not re-enter during that zone. Each data return request sent to the group by the system organizer contains the group number and the instantaneous cycle and zone numbers.

There are two types of autoresponders: global and addressed. Global auto-responses can be further divided into cyclical and continuous auto-responses. In cyclic auto-reply, the user defines the time interval during which the IPPV modules respond. In the case of a continuous automatic response, the system defines a time interval, such as 24 hours. Referring to Fig. 7, in both cyclical and continuous automatic responses, the time interval is called a zone. Each zone is assigned a single number so that the IPPV module can determine if it has already responded during a particular zone. Each zone is divided into a series of cycles. A cycle is defined as a time period

- 31 needed for the whole number of IPPV modules to try to respond. Each cycle is assigned a single number (within the zone) so that the IPPV module can determine if it has already responded during this cycle. Due to RF collisions, not all IPPV modules reach the RF receiver. To increase the probability that a certain IPPV module will reach the RF receiver, it is necessary to define a minimum number of cycles per zone. This minimum number of cycles in a zone can be changed.

Each cycle is divided into groups. The group is part of the total number of IPPV modules in the system. Each IPPV module is assigned to a specific group and is assigned a group number. The group number can be assigned to the IPPV module from an external source / user-specified / or it can be derived from a numeric address using a shift constant, as described below. Regardless of whether the assigned group number was derived, the IPPV module will only respond during its group time. Each IPPV module is further assigned a configurable re-access number. This number determines how many times the IPPV module will try to respond during its group time.

We will describe the response algorithm of the invention first in general and then in more detail.

The response algorithm of the invention is based on an effort to guarantee a constant number of response attempts. This constant is called the response rate and is measured in IPPV modules per second. The frequency of responses can be changed. The number of IPPV modules must be limited to ensure a constant response rate. We will call this constant t the maximum number of modules in a group. This maximum number of modules in a group can be changed. Based on the maximum number of modules in a group, the number of groups in a cycle can be calculated as follows:

- total number of modules_ et. pin maximum number in the group

In a system where group numbers are derived automatically from a numeric address, as shown below, the number of groups is rounded to the nearest power of 2.

The average number of modules in a group can be calculated as follows:

Wed number of modules in a group = total number of modules number of groups

We will use this number to calculate the length of the group in seconds:

- 32 Group length =^with^ · P^oiie* g⁸^^ΡΖ frequency of responses

The cycle length (in seconds) can then be calculated as follows;

Cycle length »group length x number of groups

The number of cycles in a zone can be calculated as follows:

Number of cycles - ^kopcnv^ ^ as .zone - start Zone time cycle length

If the calculated number of cycles is less than the minimum allowed number of cycles, the number of cycles is set to the minimum. The minimum zone length can then be calculated as follows;

Min. zone length = number of cycles x cycle length.

This number is compared to the zone length specified by the user in the case of cyclic auto-responses to determine if the zone length is large enough.

At the beginning of the sequence of automatic responses, the specified values ​​are calculated. The system assigns a new zone number and the initial cycle number. The auto-response control sequence is then ready to run. The system starts with the first group in the cycle of this zone and continues until the calculated number of groups in the cycle is reached. The cycle number is then incremented and a check is made to see if the total number of cycles in this zone has been exceeded / ie. whether the end of the zone has already been reached. In the negative case, the group number is reset and the sequence continues.

While the IPPV module group responds, the system receives the data and enters it into its database. After successful storage of data from any IPPV module in the database, a confirmation is sent to the IPPV module. Part of the data passed from the IPPV module to the system is the checksum of all data in this case. This checksum is an acknowledgment code and is sent back to the IPPV module in an acknowledgment message. When the acknowledgment code matches the one originally sent with the case data, the data is dropped from the IPPV module's memory. If the IPPV module does not receive an acknowledgment message from the system during the current cycle, the IPPV module will respond again during the next cycle of the current zone. When the IPPV module receives an acknowledgment message during an ongoing cycle, it will not respond until the next zone. A confirmation code is sent to all IPPV modules that responded, regardless of whether data was sent with the data. It

- 33 will reduce the number of collisions in each subsequent cycle in the zone. Addressed auto-reply or dialing is designed to select IPPV data from a specific IPPV module. The information sent to the IPPV module is the same as in the global auto-response, with the following exceptions: The numeric address of the selected IPPV module is included, the zone number is set to zero, and the rest of the information / group, cycle, feed constant, etc./ is set to that the IPPV module responds as quickly as possible, even if it has no sales report.

In the contemplated embodiment, the group size is between 2500 and 5000 terminals. Terminals are added to existing groups until all groups have 5,000 terminals each. When all groups have 5000 terminals, the number of groups is doubled so that all groups have 2500 terminals again. To illustrate, assume that the total number of P initially contains 3500 terminals in a single group. As more terminals are added to the total, the total is compared to the upper limit of 5000. When the number reaches 5000 terminals, the number of groups doubles from one to two. This creates two groups, each with 2,500 terminals. As more terminals are added, the number of terminals in both groups increases. When both groups contain 5,000 terminals each, the number of groups is again doubled to a total of four groups, each containing 2,500 terminals.

Empirically, it was determined that the optimal frequency of attempts for the considered IPPV feedback system is 50,000 attempts / hour. To keep this frequency of attempts constant, the group period must change as additional terminals are added to the system. In a given embodiment of the invention, in order to keep the frequency of attempts constant, the period of one group, i.e. the time in which each of the terminals in the group must try to send its data, must be increased from 3 to 6 minutes.

These principles can be illustrated by a simple algorithm. This algorithm can be used when groups are created automatically using the digital address bits of the terminals. Assume that at the beginning the number of groups is equal to 1 and the total number of terminals is equal to N, then if G <2 or P / G> 5000

G = 2 x G 2 / S = P / G 3 / T »K x S

Where S is the number of terminals in one group, T is equal to the period of the group and K is a constant chosen to maintain a constant number of trials, which in the example is equal to 3 minutes / 500 terminals.

The group of which a particular terminal is a member is determined using a certain number of bits in the terminal address. For example, if the number of groups is eight, the last three bits of the terminal address are used. If the number of groups is sixteen, the last four bits of the terminal address are used.

At the beginning of the group period, the system organizer sends a transaction to the IPPV processor, indicating that a new group period has begun. The system organizer then sends a global command to the terminals, indicating the beginning of the new group period and which group number was selected. The terminals include a pseudo-random number generator. This pseudo-random number generator generates a series of start times corresponding to the number of attempts and the number of reverse frequencies. E.g. if the terminal is to make three attempts and the return path uses four frequencies, the pseudo-random generator generates twelve random numbers. These random numbers are divided into the entire period of the group.

Messages from terminals to the terminal do not overlap. However, in a given embodiment, rather than generating random numbers that do not overlap within a given group period, the IPPV module will wait until the given transmission is completed before starting the second transmission, if it should start before the first message ends. It will be apparent to one skilled in the art that a set of non-overlapping random numbers can be generated and used to determine the transmission time, and the invention should not be limited in this regard.

One method of how terminals return data is one in which all terminals send the data at some point during a predetermined callback period. However, this method can result in overloading of the 2-way amplifiers and generation of undesirable phenomena in the forward path if all terminals try to transmit at the same time. Therefore, it is advantageous to divide the whole number into several groups. However, it is possible to use a group equal to the whole number of terminals.

Terminals are assigned to groups in one of two ways.

- 35 In cases where it is important that individual terminals belong to a certain group /e.g. when bridging / is required, each terminal can be assigned to a certain group using an assignment transaction addressed to the group. The cable operator may wish to assign the terminals to certain groups based on frequency. purchase or other factors associated with a particular group or subgroup of the total. There may be other reasons why cable operators allocate a given number of members to a given group, and the invention should not be limited in this regard. In this case, the number of groups is selectable in the range of 2 to 255. The group size also does not have to be the same, and the group period can be set individually to allow different group sizes. Since it is an object of the invention to eliminate bridging switching, it is desirable that the group assignment is not predetermined by the bridging switching network.

In the more general case, individual assignment to groups is not required. All terminals are controlled by a global transaction to use the lowest bits of the unique address number of the Vadres / terminal group number. In this case, the number of groups is always given by the power 2/2, 4, 8, 16, etc./. Because the lower address bits of the terminals are very evenly distributed in a large number of units, the number of terminals in all groups is practically the same and equal to the total number of terminals divided by the number of groups. The actual number of groups is determined by two factors:

The first factor is the optimal frequency R at which the terminals try to send messages to the IPPV processor regardless of the number of iterations. The second factor is the appropriate minimum period Pmin. callback. Then the total number of terminals can be divided into a maximum of 2ⁿ manageable groups by selecting the largest value of n, for which the number of terminals applies,

- = - £ R x Pmin

2ⁿ

The power 2, i.e. n, determined by this equation, then indicates the number of lower bits that each terminal must use to determine the group of which it is a member. E.g. if we specify n = 4, then there are a total of 16 groups and each of the terminals uses the lowest 4 bits of its_address as the group number.

- 36 The optimal frequency of R terminal attempts, used in the upper equation, is simply expressed as the mean number of terminals per unit time. However, each terminal has an adjustable number of retries, so the actual frequency of message attempts is equal to the number of terminals in the group times the number of transmissions each unit makes, divided by the length of the group period. During the data return period, the average frequency and length of the message transmissions made determines the message density and thus also the collision probability for any given transmission. Assuming that the mean transmission length is relatively fixed, then the frequency with which the terminals attempt to send back data is a fundamental influence that determines the probability of a collision and, conversely, the throughput of messages.

A low frequency of message attempts results in a low probability of collision, while a higher frequency of message attempts results in a correspondingly higher probability of collision for any message. However, a high success rate with a low frequency of attempts (or a low success rate with a high frequency of attempts) can result in low overall throughput. Therefore, the measure of the true success rate is the probability of success for any message times the frequency of terminal attempts. E.g. when 1000 terminals try to return data within a one-minute period and the probability that any message will be affected by a collision is 20%, then the actual success rate will be

1000 terminals x / 100 - 20 /% / min. = 800 terminals / min.

The numerically high success rate of terminals is not the final measure of IPPV throughput, as long as it does not result in almost 100% success. Because the return data represents a gain for the cable operator, all terminals must return the data stored in them. A 100% success approximation may require two or more periods in the statistical approach to return the data. To continue with the example, assume that the group has the stated success rate during the first data return cycle. 800 terminals / min. throughput may be most desirable, but it is not acceptable to leave 20% of the group without sending a message. During the next cycle of feedback data,

800 successful terminals should receive confirmation of the data received. As mentioned above, there will be no terminals that receive an acknowledgment corresponding exactly to the data stored in the secure memory

- 37 to respond again until the beginning of the new zone. Therefore, only 200 terminals that failed in the first cycle will be able to return data. The result is a much lower probability of a collision during the second cycle. To illustrate, we will assume that the probability that a message will be affected by a collision is 1%. During this one minute period, 20C x / 100 - l /% = 198 terminals will be successful. The combination of the two cycles will be an effective success rate:

/ 800 + 198 / terminals / 2min. or 499 terminals / min.

This frequency is achieved with almost 100% of the terminals transmitting the glow and is therefore a very good measure of the throughput of a real system. The optimal frequency of trials is thus defined as the frequency of trials that provides essentially 100% effective success for a given number of terminals in the shortest time interval.

The invention used a simulation method based on the IPPV data return system model to determine optimal trial frequencies. However, it should be noted that while the choice of the optimal frequency of attempts affects the performance of the system, it is not critical to the function of the invention.

The description and calculations above assume that data return is achieved to return IPPV case data from IPPV modules.

However, the RF feedback system of the present invention can be used in systems in which a number of remote units or terminals attempt to transfer stored data to a central location. Burglar alarm requirements, power management, home purchasing and other services are generally in addition to IPPV services. However, some efficiency gains can be achieved by combining the back data of these additional services with IPPV transactions, although for various transactions, especially real-time requirements such as two-way voice / telephone / transmission, it may be appropriate to use hard-coded or global commands. and answers.

For a number of reasons, such as signal-to-noise ratio and neighboring channel interference requirements, it is necessary that the output level of the IPPV transmitter carrier be set close to the optimum for the return channel. Furthermore, for low installation costs, easy maintenance, repeatability and reliability, it is highly desirable that the output level setting be as automatic as possible. The method of automatic calibration of the transmitter level is described in detail in the dependent patent. US application no. No. 07 / 498,084 of 20. March 1990; entitled Equipment for RF data transmission of cable television subscribers and the method of calibration, which is given here as a reference.

The IPPV processor communicates with the system organizer via the PS-232 serial full-duplex transmission line in half-duplex transmission format / only one direction at a time /. Any transmission format can be used, but can preferably be synchronous at 9600 Sd. This line can optionally be connected via a suitable modem if the units are far from each other. All transmitted data is preferably secured by a checksum.

The system organizer commands for the IPPV receiver include acknowledging / ACK or NAK / previous transmission from the receiver to the system organizer. When the receiver receives an ACK, it flushes its response buffer, reads a new command, and inserts a new response into its response buffer. If they have received a NAK, one of two actions will be performed depending on whether a valid command has already been received. If a valid command has already been received, then the previously entered response is simply resent, regardless of what the new command may have been. However, if a valid command has not yet been received (and thus no response is in the response buffer), the new command is read and the response buffer is full. In practice, when the system organizer detects a faulty checksum in the time check, it resends the same command with the NAK. All transmissions between the system organizer and the receiver are preferably terminated by an end of transmission indication.

Data items with several syllables are transmitted upper bits first and lower bits eventually with the following exceptions: Terminal case data and memory responses are passed unchanged. This includes a two-syllable terminal / or IPPV module checksum. In addition, a status response is also transmitted without change, which represents a memory image of important receiver parameters and data.

In this case, the more syllabic parameters are transmitted by the lower bits first and the upper bits at the end. / Intel standard format./

Checksum of system organizer or receiver /e.g. A 16-bit checksum / is generated by adding each character transmitted or received to the lowest bit of the checksum. The transfer to the highest bit of the checksum is not performed.

- 39 The result is then shifted 1 bit to the left. Initially, the checksum is not set to 0. Each message character up to / but not including the checksum is included in the checksum. The resulting checksum is inverted by the encoder and sent with the other data.

Transactions from the system organizer to the Receiver contain the following commands:

1 / Setup command - This amendment defines 4 frequencies to be used in both categories. The frequency value - 1 suppresses the use of the corresponding receiving module. With this command? the calibration parameters are also set. In response to this command, an AUTOMATIC TRANSMITTER CALIBRATION REPLY, MEMORY REQUBST REPLY, or EWJT / 7IEWING STATISTICS REPLY PACKET is sent.

2 / Initialization of a new group - This command is sent to the receiver whenever an IPPV GLOBAL CALLBACK is sent to the terminals. Informs the receiver which frequencies to tune to. It also resets a duplicate checklist. GROUP · STATISTICS REPLY 'is sent in response to this command.

3 / Query command - This command is requested by the receiver, abv sent anything that is in the queue for sending. The answer will be AUTOMATIC TRANSMITTER CALIBRATION REPLY, MEI.TORY REQUEST REPLY or EVENT / VTEWING STATISTICS REPLY PACKET. If there is no data to send in the queue, an empty EVENT / TTENING STATISTICS REPLY PACKET packet is automatically sent.

4 / Status request command - This command asks the receiver to send a listing of its current status and parameter settings. The intended purpose is diagnostics and troubleshooting.

Transactions from the receiver to the system organizer include the following:

1 / AUTOMATIC TRANSMITTER CALIBRATION REPLY - Answer The automatic calibration of the transmitter is sent to the system organizer whenever a full calibration message is received from the terminal or module. It provides a qualitative indication of the level of the received signal and the corresponding level of attenuation that was used by the terminal or module.

2 / GROUP oTATISTICS RSPLY - Answer Group statistics are translated in response to the command IKITIALIZE NEW GROUP - SDUŠtění nové

- 40 groups. Provides group statistics collected since the last command to start a new group.

3 / EVENT / VIEWING STA7ISTICS REPLY PACKST - Case statistics and tracking response packet ”- During the group period / between one command New group and another / the receiver advises statistics about programs · and monitoring from terminals or modules. The response packet provides the transmission of several program statistics and their monitoring in a single transmission format. If there is no data to send_t an empty packet is sent.

4 / MEMORY REQUEST REPLY - Answer The memory request is a listing of the memory contents of the terminal module.

5 / STATUS REQUEST REPLY - Responds A status report request is sent in response to a STATUS REQUEST command.

These commands can be further described as follows: The system organizer must send the setup command to the receiver before issuing commands for the new group. This command informs the receiver which frequencies not to tune its receiver modules to. Two frequency categories can be set, each category offering four different frequencies. A typical use of the two categories would provide a set of four frequencies for day use and a second set of frequencies for night use. Frequency selection would be performed during start-up and re-evaluated periodically.

The setting command is sent after sending the setting request in the receiver state. The status bit of the setup request is cleared when a valid setup command is received. If module D / and channel D / have a valid frequency, it is used as the frequency for SSA (signal strength analyzer). If the frequency of module D is set to 0, the frequency parameter SSA setting command is used.

The Initialize New Group command is used to mark the beginning of a group callback period. Statistics from the previous period of the group are passed to the system organizer / see answers Group statistics /. The statistics associated with the previous group period will be deleted.

The RF receiver starts collecting program statistics and monitoring responses from terminals or modules when the receiver dreams of the Initiate New Group command from the system organizer. Up to 16 duplicate messages can be received during a group callback period

- 41 from one terminal or module. However, only one of these duplicates is passed to the system organizer. All others will be canceled.

The query command requires the receiver to send any data ready to be sent to the system organizer. This response will be Automatic Transmitter Calibration, Memory Request or Case Statistics and Tracking.

The Status Report Request command asks the receiver to send a snapshot of its current status. It contains not all parameters, software control numbers, receive queue status and other status-related variables.

Answer The status of programs and watching from the terminal or module can be received by the receiver at any time. The collection of this data typically begins when a new group has been commanded by the RF receiver and global callback terminals or modules. During the global callback period, the terminal or module will transmit its program and tracking statistics up to fifteen times on four different frequencies for the return data. The receiver filters these 16 or fewer identical transmissions and passes only one of them to the system organizer.

The receiver automatically suppresses all messages that do not have a valid checksum or whose length syllable does not match the number of syllables received. The receiver will keep a record of all individual program statistics and viewing responses it receives b^&hem period of the group. This is called the accepted list. The received list contains all addresses of terminals or modules whose messages have been received. When a response arrives from the terminal, it is checked against the received list. If a matching terminal address is found, the duplicate is canceled. If the terminal address is not found, then the address of this terminal is added to the list. In this way, redundant messages are filtered or excluded before being transferred to the system organizer. The received list is emptied when the next command to initiate a new group is received. The list is large enough to accommodate the highest number of terminals that can respond during one group period.

When a response with program and tracking statistics passes the validity test and is not a duplicate message, it is placed in the message queue to be sent to the system organizer. message queue. The message queue is large enough to hold the highest number of terminals in a group if everyone sends one case. Valid messages are formed into packets for transmission to the system organizer. The auxiliary buffer, called the packet buffer, has the capacity to store the maximum number of syllables that can be transferred to the system organizer (about 2000 syllables). Messages are transferred from the message queue to the packet buffer if there is space.

The messages are deleted from the receiver's memory after the transmission is confirmed by an ACK command from the system organizer. The receiver will send program statistics and tracking packets to the system organizer shortly after the start of messages and will continue until all are transmitted. The transmission of messages remaining in the message queue to the system organizer will continue until the queue is emptied.

During the group period, the receiver will maintain management activity statistics. This is the purpose of the Group Statistics answer. The purpose is to provide the operator with feedback on both the suitability of the selected group parameters and the ability of the selected frequencies. Because the terminal or module transmits identical information on all frequencies used, the line activity statistics show whether one or more selected frequencies are to be replaced by others. The receiver maintains a number of valid responses received on each frequency. This number also includes duplicates. The receiver also determines the number of valid syllables for * each frequency. This provides essentially the same information as the number of messages, but also takes into account the variable length of the messages. At the end of the group period, the number of syllables is divided by the number of messages, so the average number of syllables per message is determined. Generally speaking, group statistics thus provide accurate data throughput on each channel and for each transmitter.

According to this information, the system organizer can automatically change the channel frequency periodically if needed due to low throughput. In an alternative embodiment of the invention, a frequency of erroneous bits or other parameters signaling low data throughput may be collected to trigger a change to another frequency. These various parameters can be monitored on the processor / receiver / display unit with four lines of twenty characters. In FIG. 14 the offer tree structure of screens for

- 43 display of monitoring, setting and calibration functions.

Group statistics are transferred to the system organizer when a command is initiated to initiate a new group. This clears all statistics from the memory. Statistics sent to the system organizer include:

1 / The total number of valid responses received on each of the four frequencies of the category during the last period of the group.

2 / Average length of responses in syllables at each of the four frequencies of the category during the last period of the group.

3 / The total number of unique answers during the last period of the group / this is the same as the number of items in the received list /.

When the system organizer begins a phase in which only addressed callback commands are sent to the terminals or modules, it must begin this phase with a new group initialization command. Although not critical, it resets the statistics from the previous group's callback.

During terminal installation and other maintenance periods, the transmitter level in each terminal or module must be adjusted so that the received level at the receiver is within acceptable limits. This is the purpose of the ATC Evaluation Response. The calibration process begins when the system organizer requests the terminal or module to send a sequence of calibration messages at predetermined attenuation levels. The terminal sends a calibration, with each of the messages containing the terminal address and the transmission level immediately followed by the calibration signal. The receiver measures the signal by comparing it with the expected level and saves the evaluation for the next signal level. The terminal then moves to the next level and sends a calibration signal again. This will continue until a complete sequence of calibration messages is sent / maximum 8 /.

When the last calibration message is received or a break occurs, the sequence will be considered complete and the ATC Evaluation response will be forwarded to the system organizer.

The calibration measurement is performed by a combination of the SSA signal strength analyzer and the selected RF receiver module, eg D. The receiver module D must be set to the calibration frequency. The frequency of module D is determined as follows:

1 / Setting to the instantaneous frequency of the group for module D, if this frequency is set to a valid frequency number.

- 44 2 / Setting to the calibration frequency of the SSA signal strength analyzer if the instantaneous group frequency for module D is 0.

3 / Suppressed if the instantaneous group frequency for module D is equal to - 1 or more than the maximum frequency number.

The calibration measurement sequence begins when the receiver receives a valid calibration response from the terminal. As soon as the end of the message is detected / the Miller code stops or is interrupted /, the pause period begins. When it expires, the measurement process begins and will continue for the duration of the measurement period. The pause period and the measurement period are determined either by the setting command or from the front panel of the RF receiver. The final signal level reading represents the average of all samples.

This section describes the operation between the terminal and the IPPV module. This sub-sequence describes the Scientific Atlanta Model 8580. When the terminal is turned on, the IPPV module performs a sequence of operations to determine the specific configuration and authorization level of the terminal. E.g. when turned on and when the IPPV module is connected to the terminal, the terminal's channel authorization data is automatically updated to include / or allow / all paid channels. In other words, simply connecting the module to the terminal may be sufficient to enable the IPPV service. Also, no bit is set in memory, indicating that RF return was used (rather than telephone or other). The module then performs a power-on calibration (hereinafter referred to as PICART) when the module has not been calibrated, to set the transmitter's carrier data output levels near the optimum for the return channel.

After the power-on reset sequence, the IPPV module begins normal companion processing. The concomitant processing generally consists of checking the instantaneous time against the stored channel monitoring times and checking the requests of the manually started calibration transmission (hereinafter referred to as MICART) from the keypad: terminal. Companion processing in the module It is controlled by a predetermined first instruction, which has a predetermined frequency, from the terminal to the module.

When switched on, the terminal reads the non-volatile memories of the terminal and copies the channel permissions, service level, debug algorithm constants, etc. to RAM. The IPPV module reads non-volatile memories and copies the group number, broadcast levels, channels of active programs, number of purchased programs, etc. to RAM. The module is then set to determine the type of terminal when receiving the next instruction from the terminal.

Upon receipt of the instruction, the IPPV module requests one syllable of data from a location in the terminal's memory to determine the type of terminal. E.g. the IPPV module would receive data determining the Scientific Atlanta 8580, phase 6 terminal type. This equipment allows the IPPV module to be compatible with a number of terminals. The IPPV module is set to read the terminal address when the next instruction is received.

Upon receipt of the instruction, the IPPV module requests four syllables of data from the terminal memory and stores the data that will come back as the terminal address. The IPPV module is then set to read the map of the allowed channels of the terminal / ie. those channels that the terminal is entitled to receive / after receiving another instruction.

Upon receipt of the instruction, the IPPV module requests sixteen syllables of data from the terminal's memory and calculates the first part of the terminal's checksum. The IPPV module is then set to read the terminal properties flags after receiving another instruction.

After receiving the instruction, the IPPV module requests one syllable from the terminal's memory and completes the terminal checksum calculation. The IPPV module is then configured to determine if a data carrier is present upon receipt of the next Instruction.

The terminal sends instructions to the IPPV module until the data carrier is present or until a predetermined power-on period has elapsed. The IPPV module then requests one syllable of data from the terminal's memory and determines whether a carrier presence flag is set. If a data carrier is present, the IPPV module reads the nonvolatile memory and determines if the module is calibrated. When the module is calibrated, the IPPV module is simply set to read the time after the next instruction is received. If the module is not calibrated, the IPPV module is set to perform PICART calibration. In both cases, the IPPV module is set to read the time after the next instruction is received.

If the data carrier is present, the IPPV module continues to check for a predetermined number of subsequent instructions (from the Predicting a predetermined time period) until the data carrier appears. If the data carrier does not appear even after a predetermined number of attempts, the IPPV module is set to read the time after receiving the next instruction and starts

- 46 normal ancillary processing, ie PICART calibration is canceled.

After the data carrier is detected, normal companion processing begins. The terminal sends an instruction to the IPPV module. The IPPV module requests four syllables of data from the terminal's memory and checks whether the instantaneous time agrees with some of the times recorded in the non-volatile memory monitoring statistics. We will explain the equipment with monitoring statistics in more detail below. The IPPV module is then set to read the terminal mode after receiving the next instruction. When a match is found between the current time and the recorded time, the terminal mode is read to determine if the terminal is on or off, so that the correct channel number is recorded. If a match is found between the current time and the recorded time, the terminal mode is read to determine if the terminal is in diagnostic mode and whether a MICART calibration was required. The step described in this paragraph will be called step G1.

If a time match is found, the terminal sends an instruction to the IPPV module. The IPPV module requests one syllable of data from the terminal's memory and checks whether the terminal is switched off or on. If the terminal is switched off, the module stores the predetermined character or characters in the non-volatile memory as the currently monitored channel. The IPPV module is then set to read the time after receiving the next instruction and repeats said step G1. If the terminal is switched on, the IPPV module is set to read the currently tuned channel after receiving the next instruction.

If a time match is found and the terminal is switched on, the terminal sends an instruction to the IPPV module. The IPPV module requests one syllable of data from the terminal's memory and stores this value in the energy-independent memory as the currently monitored channel. After receiving the next instruction, the IPPV module is set to read the time and repeats step G1.

If there is no time match, the terminal sends an instruction to the IPPV module. The IPPV module requests one syllable of data from the terminal's memory and determines if the terminal is in diagnostic mode. If the terminal is not in diagnostic mode, the IPPV module is set to read the time after receiving the next instruction and repeats said step G1. If the terminal is in diagnostic mode, the IPPV module is set to read the last key pressed after receiving the next instruction.

If the terminal is in diagnostic mode, the terminal sends an instruction to the IPPV module. The IPPV module requests one syllable of data from the terminal's memory and checks whether the correct one was pressed

- 47 key sequences. In the positive case, the module starts the MICART calibration, in the negative case it does nothing. In both cases, the IPFV module is then set to read the time after receiving the next instruction and repeats step G1.

While the sequence for the Scientific Atlanta Model 8580 terminal was detailed here, the sequence for other terminals, including those for the in-band system, is similar and will not be discussed in detail here.

The next part concerns the authorization, purchase and cancellation of IPPV cases. Unlike concomitant processing, which is based on receiving an instruction at a predetermined frequency from the terminal, IPPV case operations can occur at any time during the normal operation of the IPPV module. The terminal can accept and transfer transactions that establish or revoke authorizations to the IPPV module at any time. Similarly, the subscriber may at any time decide to purchase the show.

In this sense, IPPV operations are essentially an interruption of normal companion processing in the IPPV module.

In both out-of-band and in-band systems, terminal transactions control the establishment and revocation of authorizations. To revoke the consultation authorization, the terminal must receive the IPPV case data transaction twice. This is because the IPPV module (not the terminal) actually determines when a case is settled from transactions, and only has the opportunity to inform the terminal (via a maoy channel update request) about successful transaction transmissions from the terminal.

The basic difference between out-of-band and in-band activity is that out-of-band terminals can receive data transactions by anyone, and in-band terminals can only receive transactions on data channels. Thus, the sequence for the out-of-band Scientific Atlanta 8580 terminal will be described in detail below.

For proper processing of IPPV operations, the terminal must send an out-of-band IPPV case data transaction, which we will call an IPPV case data transaction with no higher predetermined frequency than once per second.

We will first describe the purchase of a program where the subscriber enters the IPPV channel either by direct input of a digit or by using 8 counting switches on the terminal or infrared remote control.

- 48 controls. The terminal tunes to the IPPV channel and waits for an out-of-band transaction.

When the terminal receives an out-of-band transaction, it sends the entire transaction to the IPPV module using the second instruction and determines if the IPPV module requires a channel map update. The terminal then tunes to the call channel if free time is not available, or tunes to the IPPV channel if free time is available. The terminal sends a BUT signal if the purchase window is open and this channel in the terminal's RAM is not already authorized, ie already purchased.

When the IPPV module receives an out-of-band transaction via an instruction, the IPPV module does not request a run¹ second channel map update instructions. At this point, the IPPV module performs an authorization check, which means that the specified channel is active and, if so, whether the case has already passed (different program identifications). When the case has already passed, the module submits a request to update the channel map for the next instruction, resets the active case bit for the specified channel in non-volatile memory, and changes the data format from that memory for future transmission. The procedure described in this paragraph is not called step C.

When the subscriber buys the program, the terminal sends a command after the first press of the BUT key to determine whether it is energy-independent steam! NYU full. The IPPV module responds with either the total number of stored programs or a predetermined value when steamed! NVM full. If this is steaming! full, the terminal displays.on the FUL terminal display. If not steam! full, the terminal inserts an out-of-band purchase command after the second BUY key press for the next instruction.

When the terminal receives an out-of-band transaction, the terminal sends the entire transaction to the IPPV module without using the second instruction and checks if the IPPV module requests an update of the channel map. The IPPV module then performs a second authorization check as described in step C. The terminal then sends a program purchase command to the IPPV module and receives an ACK / NAK signal from the module. In addition to the channel number, it will contain the time of purchase of the show. The terminal then tunes to the call channel when the NAK arrives, or to the IPPV channel if the ACK arrives.

When the IPPV module receives an instruction from the terminal to purchase a program, the IPPV module checks if it is steam! NVM full or whether an NVM memory phase lock loop has been affected. In the positive

- 49 case the module returns a MAK signal. Otherwise, the module is able to purchase the program and returns an ACK signal to the terminal.

When a program is purchased, the IPPV module stores the channel number, program identification / out-of-band transaction /, and purchase time in NVM memory and sets the active program flag for that program.

When the terminal receives an out-of-band transaction with a different meeting ID, the terminal sends the entire transaction using the instruction to the IPPV module and checks if the IPPV module requests an update of the channel map. The IPPV module does not require a channel map update for this transaction. The module identifies the program, revokes its authorization, and changes the program data format in the NVM memory for the next transmission. The module will include a channel mep update request for the next instruction.

These terminals also allow the purchase of an IPPV program for a video device. This is very similar to a normal IPPV purchase and will not be discussed in detail here. The basic difference is that the subscriber buys the program in advance, which causes the IPPV module to reserve space in the NVM memory for this program. This space will not be used until the show starts, but is calculated to determine if the memory is full on future purchase attempts.

The IPPV module according to the invention comprises three different types of responses: program statistics and monitoring, memory dump and calibration. The first two responses have certain properties in common, namely security data, returned to the terminal. All three responses contain the numeric address of the terminal.

Answers The statistics of programs and watching contain information concerning the number of syllables in the message, the type of message / ie. program and tracking statistics /, digital terminal address, recording times and channels that have been tuned during these) recording times and IPPV purchase data such as program identification and purchase time.

Answer The memory dump contains information about the number of syllables in the message, the type of callback / ie. memory request /, digital terminal address and information requested from memory locations.

Answer Calibration contains information about the number of syllables in the message, type of callback / ie. corresponds to the calibration /, the digital address of the terminal and the transmission level followed by the calibration curve for measuring the signal strength.

The IPPV module transmits data using Miller data encoding.

- 50 Miller coding, known as delay modulation, transmits 1 "as a signal transition in the middle of a bit interval. O has no transition unless followed by another O; in this case, a transition occurs at the end of the bit interval. Miller data coding is indicated in Figure 16.

Data transfer sequence:

Each time data is transferred, the IPPV module performs the following sequence:

A. It starts switching the transmission data line at 10 kHz. This charges the data filter.

B. Sets the gain to the minimum.

C. Switches on the switched voltage + 5 V to HF revivals.

D. Delays to about 1 ms to stabilize + 5V.

E. Sets the correct phase-locked loop frequency / read from NVM memory /.

F. Delays by about 20 ms to catch the phase lock.

G. Keying circuit to crosstalk.

K. Delays ssi by 1 ms to stabilize the final output stage.

I. Drives to the correct gain / read from NVM memory /.

J. Sends data.

When the transfer is complete, the IPPV module performs the following sequence:

A. Generates a Miller code error in the transmitted data to terminate the transmission (for the receiver).

B. Reduces gain to minimum.

C · Inserts the anti-crosstalk circuit.

D. Delays by about 1 ms to prevent frequency slippage

E. Switches off the switched voltage + 5 V.

These sequences are indicated in detail in Figure 15 using the following definitions:

Switching voltage + 5 V to the phase-locked loop, data input

Delay for capturing the phase-locked loop

Data filter charging time

Keying the circuit against crosstalk to the gain of the gain Approach to the gain Withdraw the gain

From withdrawal gain to switch-off + 5 V ^ on tLK tCHG * aB tRU tRD tQFF

One embodiment of the invention allows the system organizer to select tracking statistics regarding the channels to which a particular subscriber is tuned at predetermined times during the time period. In this implementation, the system organizer generates a global transaction that defines the four times at which the IPPV module should write to the third memory 503 / FIG. 5 / channel to which the terminal is tuned.

These times may be within a suitable time period such as day, week, every other week, and so on. To illustrate, we will assume that the system organizer instructs the IPPV module to record the tuned channel of the terminal on Sunday at 19.00, on Tuesday at 21.00, on Thursday at 22.00 in a one-week period. When the instantaneous time coincides with one of these four times, the module records the channel tuned by the terminal into the third memory 503. As mentioned, information about tracking statistics is included in the Program and Tracking Statistics response. This response contains information regarding the number of syllables in the message, the type of message, the digital address of the terminal, the recording times and channels that the terminal has been tuned to at these times, and the IPPV details of the purchase.

Although not implemented now, the system organizer could enter an addressed transaction statistics from the subscriber, who has agreed to allow monitoring of his monitoring habits.

In yet another embodiment, the system organizer may insert the addressed tracking statistics transaction into a particular group of terminals.

Fig. 8 shows in more detail the group diagram of the IPPV processor according to Figs. I and 3. The return RF signal from the terminals is transmitted in the VHP subchannel T8. The transmitted carrier can be set to a resolution of 100 kHz in the frequency band 11.8 to 17.7 MHz, which provides a choice of a maximum of 60, preferably 23 different data channels and a bandwidth of 100 kHz. The modulated carrier from the terminal or module contains BP3K 20 kbps information in the Miller code. The RF signals of the entire set of terminals in the system are combined and returned to the IPPV processor located in the terminal. The function of the IPPV processor is to receive feedback input signals, demodulate information and pass the decoded message to the system organizer.

Still referring to Fig. 8, the reverse RF signal is typically received at a single carrier level of + 12 dBmV. The IPPV processor is designed

- 52 to operate in a single carrier level of + 2 to + 22 dBmV.

Often more than one carrier frequency is received simultaneously and the total received power will be proportionally higher than + 12 dBmV. If they are on different frequencies, the IPPV processor can simultaneously receive, demodulate and decode four modulated carriers; only non-redundant decoded messages are sent from the IPPV processor control module via the RS 232 serial interface to the system organizer.

The first element of the IPPV processor, which we will describe, is the so-called input module 800. The RF return signal from the terminal is introduced from the input cable into the connector of the input module 800, which most preferably contains separate units. The input module 800 offers the input signal a nominal termination impedance of 75 0. This unit consists of a bandpass filter, a preamplifier, and a splitter circuit that divides the incoming RF signal into four RF receiver modules A through D. The bandpass filter passes the T8 band with negligible attenuation and distortion when suppressing out-of-band signals. The preamplifier compensates for filter losses and power splitting losses. RF signals are fed from the RF connectors of the input module to four RF receivers. The input module has a gain of approximately 1 dB, so the signal fed to the RF amplifiers 810 to 813 has a level of approximately +13 dBmV. All internal coaxial connections of the IPPV processor, with the exception of the input RF signal, are terminated with a nominal impedance of 50 Q. The cabling supplying + 24 VDC and ground is fed directly from the power supply (not indicated) to the input module. The input module 800 does not have a direct connection to the control module 840. All other IPPV processor receiver and synthesizer blocks are connected to the control module 840.

The second main building block of the IPPV processor is the RF receiver. The IPPV processor has four A-D blocks of RF receivers 810 to 813. These are functionally equivalent units, three of which supply 50 0 terminations at the signal strength analyzer output, so these units are interchangeable. The fourth receiver Aanal D / is indicated with a coaxial connection to the signal strength analyzer block 830. The RF receiver converts the signal coming from the input module using the output of the frequency synthesizers as a local oscillator (on the higher side). The output frequency of synthesizers can

- be between 22.5 and 28.4 MHz and is preferably between 26.2 and 28.4 MHz, which corresponds to an input frequency range of 11.8 to 17.7 MHz or preferably 15.5 to 17.7 MHz. The intermediate frequency signal is at a medium frequency of 10.7 MHz. Ceramic IF filters, tuned to 10.7 MZh, suppress adjacent channels and other mixer products and pass the desired signal. The narrowband inter-frequency signal is then detected by a circuit which provides a rough estimate of the signal strength. The output is the DC voltage RSSI, proportional to the level of the received RF signal. This voltage is fed to the control module together with other signals of the connecting wiring of the RF receivers. The RSSI information indicates the terminal's feedback signal level as received by the IPPV processor. This information is available to the system organizer.

The RSSI for a particular terminal indicates which terminals require recalibration. For this purpose, the system organizer maintains a list of too high ”too low RSSI data about the terminals, so that it is possible to include the addresses of these terminals for recalibration. This new calibration is not periodic, but is performed on a higher priority, ie on an equivalent priority with new terminals requesting calibration for the first time. Time period RSSI data tables can also be used to determine rise or fall characteristics for all 23 channels through which messages can be sent from a particular terminal. The rise or fall characteristics are then entered into the terminal so that the terminal can determine the appropriate pass levels for all Category One and Category Two channels based on the optimal results for the calibration channel.

The main function of the RF receiver is to demodulate the BPSK intermediate frequency signal 10.7 MHz. The signal is demodulated using a double balanced mixer. The demodulated data sequence is filtered and synchronized. This detected 20 kbps Miller code data is fed to the control module. The RSSI voltage derivation and BPSK demodulation functions are performed by all four RF receivers. A narrowband filtered intermediate frequency signal of 10.7 MHz at a level of approximately + 13 dBmV is introduced from the RF receiver D into the signal strength analyzer block.

A signal strength analyzer 830 is associated with the RF receiver function. The function of the signal strength analyzer block is to detect the level of the 10.7 MHz IF signal coming from the RF receiver selected

- 54 for calibration purposes. This RF receiver output is not subject to automatic AGC gain control; as a result, any change in the RF input level to the IPPV processor will result in a change in the level of the 10.7 MHz IF signal at the input of the signal strength analyzer. When the RF feedback system is calibrated, the signal strength analyzer, by detecting a 10.7 MHz IF signal, gives the control module 830 an indication of which terminal or IPPV module the output level corresponds to the received signal level + 12 dBmV. The control module 840, on the other hand, will inform the system organizer via the RS232 interface. Until the next calibration cycle (described in more detail below), the system organizer will instruct the terminal to use the level of the transmitted signal reported by the control module.

The 10.7 MHz 13 dBmV IF signal at the signal strength analyzer terminates at 50 Q. The two isolation amplifiers have a gain of approximately 30 dB for the IF frequency. The amplified intermediate frequency signal is peak detected by the diode circuit. The second diode circuit has a similar dc bias voltage. The outputs of both diode circuits add up for temperature compensation according to the known technique. The output accurately reflects the level of the IF signal, as the DC components of the diodes are compensated. This detected signal is filtered and further amplified. The final dc output signal, proportional to the level of the intermediate frequency signal, is fed to the control module.

The system organizer-controlled frequency synthesizer generates frequencies for demodulating incoming data carriers. A frequency synthesizer is a local oscillator for converting a single frequency, taking place in a RF receiver. The single frequency synthesizer block contains four separate units, i.e., synthesizers 820 to 823. The control module 840 provides frequency tuning information via serial data commands. The four frequency synthesizer units 820 to 823 are designated as frequency synthesizers A, B, 0, and D, which correspond to RF receivers 810 to 813. There are a total of 60 frequencies in the T8 channel bandwidth that cannot be set by the control module 840; however, only 23 is used according to the invention. The output frequency range is preferably 25.1 to 28.4 MHz and is transmitted down to the upper part of the T8 band, i.e. 14.4 to 17.7 MHz. The frequency resolution is 100 kHz. The output signal is at a typical level of +17 dBra.

Each frequency synthesizer unit includes an oscillation frequency divider, a PLL phase lock loop, a TG integrated circuit, and an active loop filter. These parts together form a phase-locked loop. The output frequency of the oscillator is phase and frequency coherent with a free-running 4 MHz crystal oscillator. The phase-locked loop ensures that the synthesizer output is spectrally clear and frequency accurate. The output of the oscillator excites the anti-stroke amplifier. The counter-stroke solution is used to obtain the required level + 17 dBm of the local oscillator.

The group diagram of the input module is shown in Fig. 9. The input and power division module consists of a band / preselector / filter 900. preamplifier 910. containing eg Mfflø 1134 and a dividing circuit 93Q for powering four RF receiver modules. 920.

Referring to Fig. 10, the arrangement of the IPPV processor frequency synthesizer will now be described in more detail. The frequency synthesizer assembly includes four subgroups on the printed circuit boards, as shown in Figure 10. Each subgroup is set to a specific frequency by the IPPV processor control module 840. The frequency synthesizer range is preferably between 26.2 MHz and 28.4 MHz, but may be up to 22.5 MHz to 28.4 MHz. The tuning resolution is 100 kHz. Each of the four frequency synthesizer subgroups can be set to any of 60 channels in the 22.5 to 28.4 MHz range. The RF output of the frequency synthesizer subgroup is the local oscillator signal for one of the four RF receivers in the IPPV processor. The local oscillator is on the higher side, so the RF range of 15.5 to 17.7 MHz is converted down to the receiver's 10.7 MHz IF frequency. Fig. 10 is a block diagram of a frequency synthesizer subgroup. There are again four such subgroups in the synthesizer assembly. The 1000 MHz fundamental frequency crystal 1000 is connected to a high gain feedback amplifier 1001. The amplifier is part of a large integration element with a phase-locked loop, preferably of the Motorola MC145158 type. The 4 MHz output signal is fed inside the element to a counter 1002 operating as a 1:40 frequency divider. The counter output is a reference frequency of 100 kHz, which is fed inside the element to a phase and frequency detector 1003.

The phase and frequency detector 1003 compares the two input signals

- 56 / reference 100 kHz and .variable 100 kHz / and generates error pulses when both inputs are not at the same frequency and phase. These pulses are tuned by the oscillator so that a variable signal of frequency 100 kHz is forced to have the same frequency and phase as the reference signal 100 kHz. When this happens, the output of the frequency synthesizers will be at the correct frequency. Differential error signals from the phase and frequency detector 1003 are fed from the element to the loop filter 1004 and associated components. This filter 1004 filters the error signals and converts them to an unbalanced tuning voltage which controls the oscillator 1005. The oscillator 1005 includes a transistor Q1 and associated components. Oscillator 1005 is designed so that the input tuning voltage corresponds to output frequencies in the desired output range of 22.5 to 28.4 MHz, or more preferably 26.2 to 28.4 MHz. The output of the oscillator is fed to the isolating amplifier Q2 1006. This amplifier 1006 provides a relatively high impedance and separates the oscillator from the double divider 1008 and the power amplifier Q3, Q4 1009. and N 1007 forms the total divider with the ratio = 10M + A. The counters Ν 'and A are programmed by the control module 840 of the IPPV processor by serial commands as F_ort »N_t x 0.1 MHz. E.g. the control module sets Nt to 250 for an output frequency of 25.0 MHz. it can be set by the control module to any of 60 values ​​between 225 and 284 but preferably between 251 and 284. The function of the double divider control line is to determine when the divider will divide by ten and when by eleven.

The isolating amplifier 1006 also excites the power amplifier Q3,

Q4 1009. The potentiometer setting is used (not indicated) so that the output signal level is approximately +17 dBm. A low-pass filter 1010 is connected downstream of the power amplifier, which attenuates in particular the second and third harmonics of the synthesizer output signal. The output of the frequency synthesizers is fed to the associated RF amplifier of the IPPV processor.

The group diagram of the RF receiver module is shown in Fig. 11A - C.

There are four separate RF receiver modules. Referring first to Fig. 11A, each of the RF receivers includes a mixer 1101 for converting input signals to a 10.7 MHz frequency frequency. Top injection is used. The IF signal passes through the ceramic

- 57 filters 1104 and 1105 to suppress adjacent channel signals and distortion products.

The IF signal then passes through an amplifier 1106 with a level detector 1115. The detector circuit derives a rough determination of the RSSI signal strength. The circuit of the detector 1115 is formed in a known manner, for example, from the element ΝΞ604ΑΝ. The output of the RSSI is an analog voltage, which is sent to the control module 840 for conversion to the digital form 8 sent to the system organizer.

The IF signal then passes through the directional coupler 1108. The tapped output is fed to an external output for use by the SSA signal strength analyzer module. The IF signal is then further amplified and fed to the demodulator.

According to Fig. 11B. The demodulator preferably consists of a frequency doubler 11 and an injection synchronized oscillator 1130 for carrier recovery. According to Fig. 11C, data detection is achieved via a modem filter, a clock regeneration circuit and a sampling circuit. The output of the demodulator is digital data.

Referring now to Fig. 12, a signal strength analyzer is indicated which receives a signal indicating signal strength from RF receivers. The signal strength analyzer module is used to obtain an accurate measurement of the transmitted data performance. The RF signal to be measured is fed from the intermediate frequency of one of the RF receiver modules, e.g., channel D. The signal strength analyzer module consists of a preamplifier 1200 30 dB, a level detector 1201, and a splitter 1202. The output is an analog voltage, which is sent to the control module of the IPPV processor for conversion to digital form and transmission to the system organizer. Before entering the differential amplifier 1202, two diodes are used to achieve thermal compensation, i.e., the thermal compensation diode 1204 compensates the diode of the level detector 1201.

Referring now to Fig. 13, a control module that controls the operation of an IPPV processor is indicated. The module sets up the synthesizers, monitors the signal strength, decodes the messages received by the RF receiver, checks the validity of the messages and forwards the messages to the system organizer. The control module contains a user interface / keyboard and display / for diagnostics, error reporting and configuration without switching. Fig. 14 shows the main menu from which the operator can

- 58 select the Monitoring, Setup and Calibration functions. From the monitoring menu, the operator can select six initial displays, where the SSA signal strength screen leads the operator to an RSSI (signal strength estimate). The setup and calibration menus work similarly.

According to Fig. 13, the control module board consists of six functional units: control microprocessor 1300. memory subsystem, receiver interface containing type 8097 processors and RAM memory with two inputs / outputs for each receiver, interface with system organizer and front panel interface.

The control microprocessor 1300 used in the control module is Intel 80188. It is a 16-bit processor with two direct memory access channels, four interrupt levels, three timers, 13 decoded address ranges and an 8-bit external interface.

The memory subsystem consists of dynamic RAM 1380 256 K for messages and variable storage, non-volatile RAM^- 1370 2K for parameters and sockets for 128K EPROM memory 1360 for program storage.

Two 256 K elements are used in the dynamic RAM. These are used, for example, to store group statistics, valid received messages, calibration results and other data for the system terminals.

As a result, these memories must have adequate capacity to store data packets. When messages are transmitted to the system organizer, the tables for storing messages from the terminals are reset. For each EPROM read cycle, a CAS regeneration cycle is performed before the dynamic RAM RAS. Normal instruction selections from the EPROM should be sufficient to regenerate the contents of the dynamic RAM. If there is more than 15 ps between the two EPROM accesses, the DMA controller will read from the EPROM. The 80188 microprocessor LOS memory is used to access dynamic RAM. Resetting The LOS memory must be programmed to the active memory area. After the initial setup of the DMA control unit, regeneration takes place without software influence.

There are two sockets for EPROM 1360 program memory for up to 128 K program memory. These sleeves can be used for any EPROM elements between 2764 and 27512. One sleeve is used for UCS access and the other for MGS3. After resetting, the UCS will be active in range

- 59 memories from FFBFO to FFFFF / hexadecimal /. MCS3 must be programmed to the active area.

One 2 K EEPROM memory element is available as non-volatile memory 1370 to store configuration information. The programmer must be careful not to enter the EPROM for 10 ms after recording the syllable no chip. There is no recovery delay after the read cycle. The chip is accessed from KOSO. The MCSO must be programmed to the active area.

Each RF receiver channel has its own contact processor 1310 1340 implemented by Intel 8097. Contact processors 1310 1340 decode and format Miller code data from RF receiver modules, monitor signal strength level from individual receivers and heavy signal module SSA module and control module frequency RF synthesizer.

Each of the contact processors 1310 - 1340 has its associated RAM 1311 - 1341 1 K syllables with two inputs / outputs. These memories are used to pass data and commands between the contact processors and the control microprocessor 1300. The memory contains a mechanism for bidirectional interrupts. The software can define any suitable protocol for memory usage and interrupts. EPROM memories 1312 - 1342 are available for storing programs for contact processors 1310 - 1340.

A common serial UART element 8250 is used to implement a serial interface 1350 to the system organizer. One of the interrupts of the control microprocessor I30Q is connected to the interface 1350, so that the serial interface can be controlled by the interrupt. Element 8250 can operate at frequencies up to 3θ, 4 kBd.

¹ Modem signals (RTS, DTR, etc.) are available for the answer-and-answer method. The multiplexer in the system organizer can either use or ignore these signals, as desired. The receiver has a data terminal configuration similar to a known telephone processor board.

The front panel includes a keyboard 860 and a liquid crystal display 850. The keyboard 860 preferably includes sixteen keys for decimal digits 0 through 9 and function keys such as help, next page, next line, input, reset, and menu. The keypad and display allow configuration without switching, indication of significant errors and local access to built-in test and diagnostic routines.

The liquid crystal display 850 with four lines of 20 characters is on two register outputs. The displayed data is inserted on one output and the sampling commands on the other. Display sampling is relatively slow / 1 s /.

When the key is pressed, an interrupt of the control microprocessor 1300 is generated. The coded key data is identified by reading a four-bit register. When accessing this register, the interrupt is reset. The keypad logic includes a contact bounce circuit that prevents further interruption until the end of the delay.

The control module 840 also provides power to the IPPV processor. The control module switches on the power supply of the elements as required. Each of the cables that connects this board to the RF receiver or synthesizer contains four + 12 V lines, 3 12 V lines, 3 + 5 V lines, and six ground wires.

The system organizer driver for automatic frequency selection together with the IPPV processor are responsible for automatic frequency selection for use by the transmitters of the IPPV modules associated with the terminals. The automatic frequency selection process ensures that data transmitted from the terminals to the RF processor is transmitted on frequencies with minimal disturbing noise, thus maximizing data throughput. The automatic frequency selection driver monitors the number of valid responses received by the IPPV processor on each of the frequencies. and determines the frequency of erroneous bits for each frequency. When the frequency of the error bits rises above a certain limit, the control program changes the frequency based on a certain frequency selection methodology. The process of automatic frequency selection will be described in more detail below.

Fig. 17 is a Karnaugh map showing the possible state of the return data channel, based on the number of valid responses received in the time period and the signal power level measured by the IPPV processor. The nine positions on the map correspond to nine different states into which reverse data channels can be categorized. Based on this, the classification of a certain data channel performs the appropriate function specified in a given map position.

The horizontal and vertical positions on the map correspond to the calculated ones

- 61 erroneous bit rates and signal level of the data channel. Horizontally, there are three error rate options: high, good, and low. A high error bit rate indicates that a relatively low number of valid responses has been received by a particular data channel compared to other return data channels or according to a predetermined criterion. A low error bit rate indicates that a relatively high number of valid responses has been received by the data channel. A good error bit rate indicates that an acceptable number of valid responses has been received through the data channel.

Vertically, there are three options for the signal level of the data channel: above, within, and so on. If the signal level is above the limit, it means that the transmission level used from remote terminals is too high. If the signal level is below the limit, it means that the transmission level is too low. If the signal level is within limits, it means that the given transmission level is acceptable.

Once a statistically significant number of responses have been received and counted and signal levels measured by the PSSI indicator (previously discussed), the automatic frequency selection driver determines the error bit frequency and signal level categories according to the map in Figure 17. The criteria used to measure the boundaries between the three categories for both the error bit frequency and the signal level can be preset or depend on ongoing changes in the relative operating conditions of the individual data channels used. In each case, the control program simply compares the frequency of the error bits and the level of the received signal with the respective thresholds to select the appropriate field of the Karnagh map according to Fig. 17.

Let's start with the box in the upper left corner of Fig. 17; if the error bit rate is high and the received signal level is above the limit, the control program changes the frequency of this data channel. This action is taken because the frequency of the erroneous bits is high for the instantaneous frequency, even though the level of the transmitted signal is above the limit. This means that the instantaneous frequency is very noisy and that even an increase in the level of the transmitted signal will not overcome the excessive noise that exists in this throw of the spectrum. Therefore, it is necessary to select a new frequency with lower noise.

Moving right, the next field corresponds to the condition when

62, the error rate is satisfactory, but the signal level is still above the limit. Although the inclusion of the data channel indicates that the IPPV processor received a reasonable number of valid responses, the transmission level was still above the limit. Transmission at such a level could, among other things, cause interference between data channels, and therefore the transmission level must be changed to be within the prescribed limits). Such calibration is described in detail in U.S. Pat. No. 07 / 498,084 of 20. March 1990, entitled Device for RF Data Transmission of Cable Television Subscribers and Calibration Method, which is incorporated herein by reference.

The last field primarily corresponds to the conditions when the frequency of erroneous bits is low, but the signal level is above the limit.

For the same reason, as mentioned, high-level data channel transmission is not satisfactory, and therefore in this situation, the signal level must be recalibrated to be within the prescribed limits.

The first field in the next line corresponds to a situation where the frequency of erroneous bits is high and the level of the transmitted signal is within limits. This condition indicates that the currently selected baud rate contains noise that affects the number of valid responses received by the IPPV processor. Therefore, it is necessary to change the frequency used to find another, suitable transmission frequency.

Going to the right, the next field corresponds to a situation where the frequency of the erroneous bits is satisfactory and the signal level is within limits. This is an almost ideal situation where no changes need to be made.

Therefore, no action is taken under these conditions.

The last field in this line corresponds to a situation where the error bit frequency is low and the signal level is within limits. This condition represents ideal working conditions, where almost all responses are received and the transmission level is within the prescribed limits. Under these conditions, however, no action will be taken.

The first field of the last line corresponds to conditions where the error bit rate is high and the signal level is below the limit. In this situation, the frequency may be appropriate if the level of the transmitted signal is not below the limit. Therefore, the frequency does not change, but instead the low transmission level is recalibrated as described in the dependent application, cited above as a reference.

The second field of the last line corresponds to the situation when the Frequency is

- 63 erroneous bits satisfactory, but the received signal level is below the limit. Due to the unsatisfactory transmission level, the transmission level is recalibrated again.

Finally, the third field of the last line corresponds to a situation where the frequency of erroneous bits is low, but the received signal is below the limit. As before, the transmission level must be set and therefore recalibration will be used.

Using the frequency change and recalibration procedure discussed in connection with Figure 17, the invention will eventually find a frequency and transmission level that will allow the operating conditions to be met.

The following steps describe the sequence of actions that occur during the automatic frequency selection procedure. For purposes of description, we will assume that RF receivers and IPPV modules used four frequencies for reverse transmissions. Note that this sequence is taken from the point of view of the driver for automatic frequency selection in the system organizer and is not intended as a detailed description of the functions of the IPPV module or RF processor circuits that are described elsewhere.