DK159709B - DISTRIBUTED CONTROL FOR CONNECTORS - Google Patents

Description

λ DK 159709 Β

The invention relates to a digital switching system with distributed control for use in telephone exchanges.

1 telephone switching system requires large amounts of data to control the connections, where data indicates states of subscriber and central lines connected to the switching system.

Data on couplings through the coupling matrix must also be stored, as they are required for state changes of subscriber lines and central lines. Necessary data is a subscriber eligibility, subscriber restriction, translation of telephone call numbers to machine numbers and vice versa, the central line type, and any indication of transmission routes. In a centrally controlled known system, e.g. the system EWS stores this data in a central repository which is copied and controlled by the central computer for security reasons.

When distributing the load or using several central controllers, several computers must simultaneously have access to the common storage. This causes problems in reducing the amount of data processed, which problems increase with the number of computers.

There are also known systems of distributed control and with distributed control functions. U.S. Pat.

25 3 974 343 describes a system in which the control is distributed with stored programs. U.S. Pat.

3,860,761 indicates a step-controlled switching system that uses register control and interconnects one call at a time.

30 During the development of known plants, the focus has been on achieving a high efficiency with regard to the treatment functions. Multiple treatment has been introduced to increase processing capacity, but without doing so 35 2

DK 159709 B

larger than necessary. This leads to undesirable interactions between the program packages, where changes or additions to program parts can easily lead to unforeseen collisions with other program parts during further operation.

5 When moderating program parts in known systems, it is always necessary to carry out a thorough check to determine whether the individual program parts interact improperly.

10

The main reason for this problem lies in the very design of the general control systems, where the processing function with stored program control simultaneously divides a number of tasks that depend on the incoming and outgoing traffic. In the event of a program failure or a computer error, the program may jump to unwanted storage portions, interrupting the correct operation of the software package. This is one of the disadvantages of the known systems.

20

German Patent Specification DE 2,262,581 discloses a digitally coupled PCM switching system with distributed control for use in a multi-subscriber telephone exchange with which it is possible to connect interconnections between subscribers. Thus, it relates to a closed loop system. Other telephone exchanges are not connected. Each subscriber has a subscriber coupling with an analog-to-digital converter as the interface between the subscriber's analog signals and the digital signals 30 of the connected connection. Each subscriber link allows access to a common communication path over which a number of connections are transmitted. Each subscriber coupling comprises a control unit where the control signals for building a connection are formed.

35 3

DK 159709 B

Since this known system relates to a loop line, it is only connected to subscribers, while there is no connection line to other telephone exchanges, there will be no different transition conditions for controlling the individual connections. The system does not include a coupling matrix, since the control process confines itself to connecting both subscribers to the loop line each time. The subscriber link is uniformly expanded, which is why a firmly embedded control program is sufficient to produce all conceivable connections. The testing processes are similarly simple. Each free time channel can be used for testing. Conflicting situations when testing and modifying different drivers can therefore not occur.

15

Against this, the technical assignment position is broader in the purpose of the application. Control of a switching system that is connected to an external network is required.

Only here is the conflict situation formed. A central control-20 program must ensure that program packages moderated for certain connection types remain compatible with the remaining program packages. A dual network of roadways can be a very costly solution to the problems. A test of the compatibility of all the program packages 25 in a common control network cannot be performed when the changes occur during operation.

This object is achieved by the telephone switching network being characterized in that the common communication paths are connected to a switching matrix, digital switching units are connected to the switching matrix as interface circuits for central lines and subscriber carrier system lines, digital control signals and voice signals are generated in the digital switching units. 35 analog wires are converted into digitized voice

DK 159709 B

4 signals using analog-to-digital converters and the digitized voice signals and digital control signals for building a selected connection are transmitted in uniform channels through the common communication paths and in the switching matrix, which digital control signals individually control connections between subscriber lines, respectively. . subscriber wires and digital switching devices.

The switching system according to the invention is distinguished by the fact that there is no mutual influence between changed program packages, since each connection is connected by means of its own control. In addition, it is possible to test new program services in operation in limited settings during the program changes. Since the controllers in the subscriber couplings and in the digital end surfaces do not exchange programs, minor differences in the different coupling types are respected.

Suitable details of the invention are set forth in claims 2 to 7.

The invention will be explained in more detail below with reference to the drawing, in which: Figure 1 shows a block diagram of a preferred embodiment of a distributed control switching system according to the invention; Figure 2 shows a diagram of the cost as a function of the number of connections, Figure 3 shows a block diagram of a group of subscriber links with a common storage, 35 5

DK 159709 B

Figure 4 shows a block diagram of a common store, Figure 5 shows a block diagram of a subscriber coupling with a call controller, and Figure 5 shows a block diagram of a translator.

A coupling matrix 102 shown in Figure 1 performs a central function in a coupling system. The coupling matrix 102 10 must be of a non-blocking nature. Coupling matrix 102 may be a concentrator or deconcentrator or other type of PCM coupling matrix, which enables time and space coupling, so that it is possible to transfer information from any time channel on an incoming multiplex line to another random time channel on a second outbound multiplex line. The switching matrix 102 comprises an internal path selection control for controlling traffic through the switching internal communication paths, thereby allowing a distributed control of the connection of connections by the subscriber. Diagnostic programs for locating errors down to the level of an interchangeable element (hardware elements or modules) are decentralized and include a microprocessor control of the subscriber lines. A security block 25 can contain between 1 and 60 wires, and each security block is equipped with a microprocessor. Decentralized diagnostic programming serves to suppress mutual interaction between fault areas on one line and traffic on other lines. By transferring the diagnostic program to the individual microprocessors, their performance need not be optimized, since the distributed programs can be designed so that testing can be performed at each stage. The coupling matrix 102 is shown simplified and comprises a first coupling step with the coupling subgroups 1, 2, 3 and N with reference numerals 6

DK 159709 B

104, 106, 108 and 110, respectively. Each coupling subset in this step comprises a road search controller designated by reference numbers 112, 114, 116 and 118. The present coupling step in coupling matrix 5 102 is shown only with coupling subset 1 and N, denoted by 120 and 122, respectively, and associated controls 124 and 126.

An input to the switch matrix 102 is connected to a multiplexer 148 through a multiplex line 204, and the multiplexer 148 interacts with a group of subscriber couplings 128. From this subscriber coupling, analog two-wire couplings 132 to 134 extend to the subscribers. In the subscriber switch 128, a microprocessor ensures that the traffic coming from the subscribers analog-to-digital is converted, while the reverse traffic to the subscribers digital-to-analog is converted. For example, a microprocessor 8080 or the like may. serve as a microprocessor in the subscriber coupling 128 serving the subscriber wires. The construction of the subscriber link is the subject of another patent application. Digital subscriber lines 130, subscriber carrier system lines 136, and digital center lines 138 are connected to a digital switching unit 140 which performs a storage and controls the calls, and is connected to the switching matrix 102. Required databases (translators, etc.) 142 and other stores 144 (for fee counting , monitoring, etc.) and service circuits 146 are connected to the switching matrix 102. The translator in the database 142 interprets the subscriber's rotated or keyed 30 digits, which corresponds to the function of normal translators.

They are also used here to realize the distributed control, being coupled with the only kind of connection path formed between the subscriber coupling 128 and the coupling matrix 102, one of these communication paths being designated 204.

DK 159709 B

the translators are reviewed in more detail in connection with Figure 6. The same switching matrix 102 provides both a data communication path and a voice path between the subscriber lines. Since the individual subscriber couplings 128 5 themselves control the construction of a path to the coupling matrix, the functions required for a central processor fall away. The subgroups 108 and 110 of the coupling matrix 102 are connected in the same manner as reviewed for the subgroups 104 and 106.

10

Figure 2 shows different curves and the cost per unit. wire as a function of the number of wires connected for different systems. Curve a represents the well-known electromechanical step-by-step switchgear where control is performed directly by the subscriber by means of the turntable. Such a system can be expanded over a large area with moderately rising costs, which means a lack of flexibility for the system when it reaches a certain size. The curve b represents a register / over-20 controller controlled stepwise system, in which register transmitter and translator functions are used to make the numbering plan larger and add greater performance. Duplication of common parts for the register transmitter and translator function is necessary, which is why the cost 25 per Wiring in the plant also increases the fewer wires it contains. Curve c represents known control systems with wire routing logic, such as No. 5 Crossbar system.

Such a system not only suffers from the drawback already mentioned where parts must be duplicated, but in addition it can only be expanded within a relatively small area, e.g. 8-multiplied. The control with a common wiring logic also does not allow direct subscriber control with coupling via the communication path. The curve c also shows the cost per unit. line for electronic switchgear 35 with central control through stored programs. Here too

DK 159709 B

development opportunities are limited due to processor performance.

p

The curve d shows the cost per unit. line of system i-5 according to the invention. Since the control elements, such as are microprocessors, are mounted per. line or per. line group, and since the system is modular (which has not yet been reviewed) and can be manufactured uniformly, and comprising a standard switching unit for the switching matrix 10 instead of transmitting control signals through signal channels utilized by the prior art, the system of the invention can easily expanded, which is why the cost per line does not vary much from 1000 subscribers to 100000 subscribers. The total expansion potential equals at least one thousandfold. As the switchgear is expanded, the throughput capacity automatically increases, along with this, a modular expansion of the switchgear raises the upper limit of expansion compared to what is normally found in ordinary control and main centers with stored program without any loss of flexibility. The module structure also enables the addition of new features and services to one or more modules, without this requiring extensive testing of existing features, which is precisely a limitation of known control systems with stored programs.

The call control, the common storage and the distributed control are described with reference to Figures 3, 4 and 5.

30 To control the calls, each subscriber switch 128 (Figure 5) includes a call controller 302 which controls incoming and outgoing call halves at different times. The call controller 302 comprises a microprocessor 402 with its own memory 516, an interface circuit 512 to 35 a common program memory, an interface circuit 518 to a 9

DK 159709 B

power supply unit, two multiplex connection lines 212 and 214 for the switching matrix, also common to other subscribers. The call controller 302 has an address capacity of 256 kbytes and allows a digital filtering.

5 The call controller 302 enables direct and alternating current control up to 300 Hz to generate a supply voltage and ring current functions while the speech signal processing is in the range of 300 to 3800 Hz. The speech signal processing is performed with a speech frequency processor controlled by microprocessor 402. Each two-wire line 132 is connected to an analog-to-digital converter 502 and a digital-to-analog converter 504 through a protection circuit 510. The digital output signal from the analog-to-digital converter 502 is digitally filtered in processor 500 and converted to a sequence of bits, e.g. in a linear 14 bit PCM serial code with extra bits for controlling the switch matrix 102 and for information transfer between different connection controllers and the translator. The digital filtering 20 also performs a two-wire to four-wire transmission which compensates for attenuation characteristics of the given line 132. Microprocessor 402 is programmable to be used for balancing and loss and gain control, which is equivalent to a resistive attenuation link. . Further, the output signals from the analog-to-digital converters 502 are filtered for tone detection. The processor 500 also generates digital signals and transmits them to the digital-to-analog converter which then emits an audible 30 tone in the range of 300 to 3800 Hz to signal a busy tone, ring tone, etc. to the subscriber line.

Number pulses or corresponding tone signals are received and processed by microprocessor 402 which determines whether it is necessary to provide access to the common databases and

DK 159709 B

10 the translator for further data processing. A set of instructions pertaining to the operation of the individual wire is available from a common store 200 (Figure 4) through a data bus 306 and an interface circuit 512 5 (Figure 5). Such access is limited to a group of subscriber links or a data bus and thus does not use the switch matrix 102 to obtain this data. Thus, the program control instructions are distributed on the individual wires, so that the various wiring blocks may have different combinations of programs representing different service classes or properties. Thus, it is not necessary to store all program instructions on a distributed basis, thus saving storage space. Interaction across the multiplex lines 212 and 214 of various combinations of program control instructions is prevented by the switching matrix 102. This creates the possibility of changing capacity, as well as adding or removing program bits. A single microprocessor also has only 20 access to the part of the programs that belong to the incoming or outgoing half of a connection, which of course depends on its direction. The interface circuit 512 also contains a clock line 506, a data bus 306 for data and addresses, an instruction line 308, and a main clock line 514, which enables the above-described information exchange between the modules. The interface circuit for the switching matrix 502 comprises two output circuits 520 and 522 as well as input circuits 524 and 526.

30

The distributed control provided by this system, with which each connected connection utilizes a single-use microprocessor throughout the call, eliminates previous need for complicated microprocessor splitting algorithms 11

DK 159709 B

between multiple calls. The distributed control can be provided by one microprocessor per second. Erlang, pr. adaptation or per. security block. A microprocessor is connected to a cord for at least the time that a conversation is held on that cord. In the preferred embodiment, each wire is provided with one microprocessor.

Figure 3 shows an arrangement of sixty subscriber couplings 10 128, which can be considered as a group for a multiplexer 148. Each subscriber coupler 128 forms an interface between the subscriber line 132 and the coupling matrix 102. Here, the two-wire communication is converted to the four-wire communication, 15 digital signals are filtered, digital voice signals are processed and call management is performed. The proprietary memory microprocessor in the subscriber coupler performs call control, translation, and generates control signals for path selection as well as various other 20 diagnostic functions and is connected to the common memory 200 through the interface circuit 512 (Figure 5) of the subscriber coupling 128. group of sixty subscriber couplings 128 uses common store 200. Each wire such as wire 212 of subscriber coupler 128 is multiplexed to a group 204 of 32 channels, i.e. transmitted through a common communication path transmitting a linear 14 bit PCM code with a sampling frequency of 8 kHz. Two channels are also intended for information exchange with other 30 modules in the system, ie for synchronization and for signal purposes. Another operating unit 216 is connected.

Each microprocessor 402 (Figure 5) has its own storage 516 with a storage capacity of 4 kbytes, and has access to a common storage 200 (Figure 3) with an address capacity of 256 12

DK 159709 B

kbytes.

Both the program repository and the fixed data repository are in common use. However, there are 5 stocks attached to the microprocessor and the storage with start instructions for the microprocessor is not common, which is to prevent interaction between the microprocessors. Each multi-processor controlled storage system presents problems with a processor's access to the storage, the required access time for certain sections of the common storage, as well as the complexity of hardware and software to combat said problems.

The common memory may consist of a multiple-in-one memory store where each microprocessor 402 (FIG. 5) of a dialer controller 302 has access through its own interface circuit 512 and an associated data bus 306 connected to other data buses for other devices in node 318 (Figure 4) and continues as a common 20 data bus 310. Each interface circuit 512 is further provided with an individual instruction line 308. As already mentioned, a maximum of sixty subscriber links have access to the common store. A decision circuit 316 ensures that only one microprocessor at a time has access to the common storage 200, thereby avoiding the hassles of simultaneous work in the storage. A control circuit 312 controls the addressing in the common storage 200 through line 320, and where data is transmitted over line 322 and over bi-directional data buses 310 and 306 30 to interface circuit 512. To ensure that data and address transfer are in order, verify parity bits in interface circuit 512 and control circuit 312. Common memory 200 may be a semiconductor circuit with a RAM memory where 32 bits of data words are organized to constitute the address mentioned 13.

DK 159709 B

256 kbyte capacity. A master clock signal transmitted from the main clock line 514 is used in the clock distribution circuit 314 to provide various clock signals transmitted through the channel clock line 506 5 and the interface circuit 512, as well as over corresponding wires to the decision circuit 316, as well as to the storage control circuit 200.

The translator 202 (Figure 3), the sixty subscriber circuits 128 and 10, the common storage 200 are connected to the coupling matrix 102 over the multiplex lines 204, 206, 208 and 210, respectively. The line 204 leads to an input X of a subset Y and the line 208 to the input X + 1 on the same subgroup. The conduit 206 leads to the input X of 15 a subgroup Y + 1 and the conduit 210 to the input X + 1 of the same subgroup Y + 1. As already mentioned, each of the conduits 206 to 210 comprises a multiplex conduit of 32 channels. Each subscriber connection 128 is connected to two multiplex wires with 30 voice channels each.

In addition, synchronization channels are also connected so that the outputs of the subscriber connections and the inputs of the switch matrix 102 are synchronized to provide the necessary parallel-to-serial or serial-to-parallel conversion. The microprocessor 402 storage 25 5 1 6 (FIG. 5) may comprise a masked ROM or PROM storage. The storage 516 has a storage range of the order of 4 to 8 kbytes, which area can be overwritten for firmly embedded software and variable data which may contain data on the degree of service.

The translator 202 of Figure 6 can advantageously be used in an organization with distributed control and operates over the only data path between the group of subscriber links and the other subsystems of the switching system which provide 14

DK 159709 B

is brought about by the switching matrix 102. The inadequacies of known translators in the switching systems' centrally stored programs for data changes in the telephone exchange or for the subscribers are overcome by the switching system according to the invention. The translator 202 contains a memory 550, a control processor 552 with its own microprocessor 554, and a corresponding program memory 556 and eight access ports to the switch matrix 102, where the access ports 558, 560 and 562 are shown. The translator can be duplicated if required in conjunction with increased traffic, reliability or longevity.

The translator receives correction information of various kinds, capacity data, statistical information 15, etc., while the normal translator functions are performed: ie. translating subscriber number into position number using a table in the storage 550; translation of position number to subscriber number; translating area code or central code into a central line; and translating a central line into an equipment number.

Each subscriber connection 128 (Figure 3) is, as already described, connected to two multiplex wires with 32 channels each. Each of these multiplex wires is connected to an input to switching matrix 102. A typical telephone exchange is constructed with 960 subscriber connections identical to subscriber switch 128 and connected to switching matrix 102. The 30 subscriber couplings connected to the multiplex line have a different output over which they are connected to another multiplex line. The two multiplex wires serving 30 subscriber connections are connected to two corresponding inputs on two consecutive subgroups 35 in step 1. The 60 subscriber connections having a common 15

DK 159709 B

In stock, there is also 1 connection with four multiplex lines. Each subgroup in the first step has four inputs, so the 960 subscriber links mentioned must be supplied to eight subgroups in the first step. Each subscriber connection 5 128 has two equipment numbers through the connection to two multiplex wires, where the converter continuously monitors the state, ie. busy / free, in each outgoing and each incoming call half in the subscriber link 128. The answer to a query for translating a subscriber number to an extension number contains an available position in the number and an indication of whether the apparatus associated with the other position, is free, busy, or ordered. When both subscriber connections are recorded, this information is returned to microprocessor 552 15 (Figure 6). The microprocessor which controls the outgoing call half selects a path through the network to the subscriber link characterized by the position number and outputs the necessary information to make a call. The microprocessor which controls the incoming call half sends a confirmed session signal to the translator, confirming that the subscriber link is currently busy and to identify the equipment that initiated the call.

Translator memory 550 may comprise a CCD memory or magnetic bubble memory or other fixed memory capable of containing at least 90,000 words, 80,000 of which are used for translator memory, while the last 10,000 is used for the translator program's memory 30. The word length, depending on the data structure, can take a length of 16, 24 or 32 bits. The access ports 558 to 562 in the direction of the switching matrix are identical to the subscriber couplings and can be identified and selected by means of the position numbers in the same way as in the subscriber couplings. The distribution of 16

DK 159709 B

the access ports are such that one output of a switch module does not block more than one access port and that the output of a subset in a loop does not block more than half of the access ports. The 5 position numbers selected for the access ports are selected so that an algorithm 1 program stored 556 can derive another position number from a given position number. Functionally, each access port contains means for receiving messages from a multiplex line channel defined by 10 Sit equipment number as well as means for identifying control messages sent between the microprocessors on the channel. Further, each access connection includes a storage register where one or more such messages are temporarily stored, an output layer, and means for passing such messages to the correct channel on multiplex lines 564, 566 and 568, and means for sending messages to keep the transmission paths vacant, while the translator microprocessor 554 generates output data.

20

Access ports 558, 560 and 562 also include access lines. Data is extracted from memory 550 in response to the messages received through the access lines. Data changes in the memory 550 are controlled by the microprocessor 25 depending on the program in the program memory 556. The microprocessor 554 may be similar to the microprocessor used in the subscriber circuit 128 and the microprocessor 402 shown in Figure 5 which is part of the call controller 302 which is again part of the subscriber link. The translator's warehouse 550 contains the necessary translation tables. The probability that the access time for a translation exceeds 4 milliseconds is less than 0.001. The average value for the time a complete translation takes is less than 2 35 milliseconds. Data changes for subscribers and for 17

DK 159709 B

the switchboard consists of reprogramming the bearings 556, thereby providing additional customer data or an extension of the number of subscriber lines or central lines. The translator is supplied with two additional 5 clock channels X and Y over lines 570 and 572.

10 15 20 25 30 35

Claims (7)

A distributed control PCM switching system for use in a telephone exchange with a plurality of subscribers, whereby it establishes a requested connection between subscribers, with which a subscriber coupling comprises an analog-to-digital converter as interface circuit between the subscriber's analog signals and the digital signals in 10, the switched-on connection with which each subscriber switch translates allows access to common communication paths, on which a number of connections are established, and with which each subscriber switch translates a control unit in which control signals for establishing the connection are generated, characterized in that the common communication paths (204) is connected to a switching matrix (102), that digital switching units (140) are connected to the switching matrix (102) as interface circuits for central lines (138) 20 and subscriber carrier system lines (136), that digital control signals are generated in the digital switching units (140) and speech signals fr a connected analog wires (136) are converted to digitized voice signals using analog-to-digital converters, and that the digitized voice signals and digital control signals for building a selected connection are transmitted in uniform channels through the common communication paths (204) and in the switching matrix (102), which digital control signals individually control connections between subscriber lines 30 (132), respectively. between subscriber wires (132) and digital switching devices (140). Switching system according to claim 1, characterized in that each subscriber line (132) has an interface device (128). DK 159709 B Switching system according to claim 2, characterized in that each call control unit (302) comprises a microprocessor (402) with an associated storage (516). Switching system according to claim 3, characterized in that a group (148) of interface devices (128) is associated with a common storage (200) accessible through the call controller (302). Switching system according to claim 4, characterized in that each group (148) of interface devices (128) is associated with a translator accessible through the common communication path (204). Switching system according to claim 1, characterized in that each call control unit (302) controls only one half of the call in the connection of a connection. Switching system according to claim 1, characterized in that each common communication path (204) is a time-multiplexed line.