CH623948A5 - Data processing system with a multiplicity of on-line stations and use of the system for recording details on manually established telephone connections. - Google Patents

Abstract

In order to simplify the charging and accounting procedures in a telephone network and, at the same time, to carry these out in a more reliable manner, the system exhibits a multiplicity of switching devices (11, 12, 18; 21, 22, 28; 31, 32, 38) with keyboards and display devices. These are connected group by group to control devices (10, 20, 30) for the user which respond to commands keyed in in order to obtain data from peripheral storage devices and to record details of the telephone connections on magnetic tape. Each control and peripheral device exhibits microprocessors for controlling the individual functions and carrying out tests which determine the operating states of the devices and generate relevant display signals. All these devices are connected to a monitor (6) for recording these display signals. Each device can exhibit a monitoring clock generator with a counter. The latter is regularly incremented and periodically set to zero, responding to a software command. Thus, when a central processing device enters a software loop, the counter exceeds a threshold value and generates a signal indicating the presence of such a loop. The same timing generator can also reset the central processing device into a starting position and delete all data recorded as consequence of a program execution up to this time. Finally, there can be a diagnostics device so that a technician can carry out routine checks of the individual devices. <IMAGE>

Description

The object of the invention is to provide a data processing system of the type mentioned at the outset which does not have the disadvantages of known designs. This object is achieved by the measures defined in the characterizing part of patent claim 1.

Particularly advantageous refinements of the data processing system are described in claims 2 to 11.

Preferred embodiments of the subject matter of the invention are described in more detail below with reference to the drawings, which show schematically and with the omission of components which are insignificant in this context:

1 shows a data processing system for recording details of manually made telephone connections;

FIG. 2 shows a detail from FIG. 1 with more details;

Fig. 3 shows a fault monitoring unit of the system of Fig. 1;

FIGS. 4 and 5 two different peripheral devices of the system of FIG. 1;

Fig. 6 is a monitor device of the system of Fig. 1;

FIG. 7 is an illustration for explaining how monitors of central processing devices (hereinafter referred to as CPUs) are coupled to the monitor device of FIG. 6; and

Fig. 8 is a switching circuit.

The data processing system of FIG. 1 has data input and evaluation means such as, for example, a large number of switching devices 11, 12, 18; 21,22,28; 31,32,38 located at the operator's workplaces. Each switching device (hereinafter referred to as OPE for short) is intended for an operator and is connected to a control unit (hereinafter referred to as OCU) 10, 20, 30. For explanation purposes, the latter is designed such that it receives information from eight OPEs, divided into two groups of four OPEs each, and that it can transmit information to the eight OPEs. Furthermore, the OCUs 10, 20, 30 are connected via a telecommunication cable A, B to a multiplicity of data handling and / or storage units, such as, for example, a peripheral main data store in the form of a

Magnetic tape cartridge recording device 3A, 3B, or magnetic tape cassette recording device with a peripheral tariff and credit card checking device 4A, 4B and with a peripheral step change and code conversion device (national-local) 5A, 5B. It should be particularly pointed out that here the telecommunication cable A, B, as well as each of the peripheral devices 3A, 3B; 4A, 4B; 5A, 5B are present in duplicate and that the circuit is designed in such a way that if one duplicate fails, the other duplicate continues to operate the system. In addition, the system is designed in such a way that operation of the system is fully guaranteed even if one OCU fails, because then two other OCUs automatically take over the functions of the failed OCU.

In addition, all OCUs 10, 20, 30 are connected to a monitor device 6, which in turn is connected to the peripheral devices 3A, 3B; 4A, 4B; 5A, 5B is connected to receive signals to indicate their functional state. The monitor device 6 has devices for printing out details about any abnormality in the operation of the system,

which the monitor device 6 during periodic sampling of the outputs of the OCUs 10, 20, 30 and the peripheral devices 3A, 3B; 4A, 4B; 5A, 5B. Furthermore, a fault monitoring unit 7 is provided, which is used for receiving information (recorded in the monitor device 6), for operating a fault monitoring connection unit 8, for applying test signals via a fault monitoring cable 9 to the OCUs 10, 20, 30 and the peripheral devices 3A , 3B; 4A, 4B; 5A, 5B and for receiving response signals over the same cable, for the purpose of assisting a maintenance person who has to find the cause of any errors that may occur.

The individual components of an OPE are not shown and need not all be localized at the operator's workplace; they include manually operated buttons and a display unit (hereinafter referred to as VDU) for the delivery of visual information. The necessary logic and connection circuits can be available at the workplace or in the relevant OCU 10, 20, 30. The keypads can be arranged completely separate from the buttons and switches for the operator's control of the switching operation; however, some interconnection between these components is preferred in order to ensure that the exact time between the establishment of a connection and the disconnection thereof is transmitted to the OCU 10, 20, 30, possibly with details of the subscriber station to be called and the facilities required by the subscriber. Each OCU 10, 20, 30 is configured to transmit the information required by the operator to the display unit, possibly also illuminating certain keys in order to provide some required information, for confirming the actuation of certain keys and / or for that check correct operation of the components of the relevant OPE.

As mentioned, each OCU 10, 20, 30 is connected to control eight OPEs, but the OCUs 10, 20, 30 are designed in such a way that they are able to process information from twelve OPEs and also transmit them to twelve OPEs. Each OCU 10, 20, 30 has two groups of four buffers, each buffer being accessible for information from and to a corresponding OPE and two spare parts, each for processing the input signals from four further OPEs or for transmitting a group of four OPEs on an adjacent OCU 10,20,30. In the event of failure of an OCU 10, 20, 30 (if it is discovered by testing performed internally by each OCU 10, 20, 30), the eight OPEs associated with that OCU 10, 20, 30 will become in divided into two groups

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

623948

two adjacent OCUs 10,20,30 transmitted. Each OCU 10, 20, 30 contains a microprocessor, for example of the "INTEL 8080A" type. As is well known, a microprocessor is similar to a small computer in which the necessary data processing is effected by a single integrated circuit. Other types of microprocessors can be used if they meet the requirements set by the OCUs. The necessary power supplies, clock pulse generators, data, program and buffer memories are connected to each microprocessor.

As shown in Fig. 1, each peripheral device 3A, 3B; 4A, 4B; 5A, 5B duplicate and each duplicate includes a similar microprocessor as used in the OCUs. The peripheral device 4A, 4B stores the information regarding the tariff levels of the various possible telephone connections and other services such as the delivery of personal calls. Since the tariff level also depends on the time of day at which the connection is established, this device 4A, 4B also comprises independent clocks, not shown, which are driven by so-called main timers and transmit this time, so that the correct tariff level is specified for the connection and that OCU 10,20,20 is supplied. Devices not shown are used to reset the clocks under the control of the maintenance person. Since certain subscribers make use of the possibility of charging telephone charges to a credit card account, the operator must be able to check the validity of the credit card number given by the caller; for this reason the tariff and credit card checking device 4A, 4B is able to check the number and to indicate to the operator whether the number given is valid or invalid. The latter can also provide other information, for example if a certain account has not been renewed.

The clocks in the two duplicates are independent of one another, but are compared with each other every 30 minutes, and if the error is greater than 30 s, an error signal is transmitted to the monitor device 6 so that the maintenance person is informed. Upon receipt of such a message, the maintenance person uses the error monitoring unit 7 to set both clocks to the time indicated by a mother clock.

The peripheral device 5A, 5B provides the operators with step change and national code / local code conversion information, in relation to the area codes used for certain operations or connections, whereby the OCU 10, 20, 30 is transmitted the distance factor of the connection to be established that the appropriate load level is used.

The most important peripheral device, the main data store 3A, 3B, records on two separate magnetic tapes details of the manually made telephone connections, together with the identification of the subscriber, which is given, for example, by his connection number or credit card number. The duration of the connection, the time of day, the tariff level and all other details are also recorded on the magnetic tapes if, for example, it is a personal call. When a cassette has been filled or after a certain period of time, the cassette is replaced; The filled cassette is sent to central accounting, where the charges for the connections are calculated and debited to the account of the subscriber who is subject to the fee.

As mentioned, the communication cable A, B, with which the peripheral devices are connected to the OCUs 10, 20, 30, is duplicated in order to ensure that a single fault cannot shut down an entire section of the system. In addition, it is possible that a faulty duplicate can be checked and repaired without interrupting the operation of the system as far as the operators are concerned. In order that the maintenance personnel can quickly locate and repair faults in any section of the system, the monitoring device 6 is provided, which is connected to each OCU 10, 20, 30 and to each individual duplicate of the peripheral devices via a system of independent cables that make it possible to estimate the functioning of each test item in relation to relevant parameters, such as the filling level of buffer memories. In addition, all microprocessors are programmed in such a way that they check their intrinsic function and transmit a message that everything is in order or that there is an error, which messages periodically reach the monitor device 6, for example every four seconds. The monitor device 6 has these messages printed by means of a remote printer 6A in order to check the overall condition of the system and also to obtain statistical information which enables decisions to be made as to whether, for example, additional operating personnel, additional peripheral devices or additional duplicates are required.

If the monitor device 6 is not able to precisely identify the faulty assembly, a maintenance person is able to use a fault monitoring unit 7 which contains a further microprocessor and a large number of fault checking instructions on test cassettes which are stored in the fault monitoring unit 7 on Input 7A will be introduced. The fault monitoring unit 7 is designed in such a way that it detects the presence of a fault at any point and locates the latter, in such a way that it can also be operated by non-specialized operators who have no training whatsoever with regard to the assemblies used in the system. In order to carry out the error check, the error monitoring unit 7 triggers a test signal which, via the error monitoring cable 9 and via the error monitoring connection unit 8, reaches the individual OCUs 10, 20, 30 or the individual duplicates of the peripheral devices separately; Corresponding output signals are transmitted back to the fault monitoring unit 7 via the same fault monitoring cable 9 and the fault monitoring connection unit 8. The remote printer 6A enables the maintenance person to enter commands and data via his keypad and prints the results. For this purpose, the remote printer 6A is connected to the monitor device 6 via a line V24; a second line connects the monitor device 6 to the error monitoring unit 7.

Another function of the error monitoring unit 7 is that it makes it possible to transfer data into the peripheral device 4A, 4B, since it is likely that at least the credit card data will change frequently.

The CPUs 122, 307, 406 can fail during operation and are therefore programmed to carry out periodic checks in order to monitor their operating state and that of the units or devices which they control. To achieve this, each CPU 122, 304, 406 comprises a monitoring clock, consisting of a counter, to which a pulse with a pulse frequency of 1 kHz is applied. The programs of the microprocessor, which is present in the CPU, are designed so that they control the counter at intervals that do not exceed a quarter of a second and reset it to the zero value. If the counter is not activated during this time, for example due to the error of a program to remove the microprocessor from the command signal loop or due to a material error in the microprocessor, the uncontrolled counter starts an "expiry time" to allow another time period within which the Counter can be controlled. If too

4th

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

5 623948

within this expiry time the counter is not controlled, they are recorded. If the data in the bei-so the program of the microprocessor is interrupted and the FIFOs differ, then each of them is checked for validity in a simple way back to a reference position in the program sequence, for example by means of a Parisetz. If a program interruption occurs, a crime review; if a set of data is incorrect, then who signal is generated and this signal is transmitted to the monitor device 6 5 in both duplicates only the respectively valid data so that it records details of the error. If, on the other hand, the data differ, but both can. A second error when driving the counter sets of data appear to be valid, then each duplicate within the quarter-second interval leads to the 3A, 3B being designed such that it detects the device or duplicate relating to its own FIFO as failed ; stored data, but to monitor device 6 at this time, if the device is an OCU 10, 20, 30, 10, an indication is transmitted that the two sets of data differ from one another in the switching of the relevant OPEs to the neighboring ones. It is also possible for any two OCUs to be marked. ner type is inserted at this point in the record to

As mentioned, each OCU 10, 20, 30 receives input indications that the mark immediately preceding signals from eight OPEs, but has inputs for twelve OPEs. the data or immediately following data are suspect. As a result, if an OCU 10, 20, 30 fails, as already indicated, the system, even after the OPE normally assigned to it, is kept in full operation as two groups of an OCU failure.

four OPEs to corresponding other OCUs 10, 20, 30- For this purpose, each OCU has a sufficient number of input benches, which then work with twelve OPEs each. Output channels for one and a half group of OPEs

The inputs from the OPEs to the OCUs 10, 20, 30 are (Fig. 2). Each OCU is transmitted on its channels by means of a first "in first - 20 coupling device 111 to 114, 119, 129, assigned in groups of four OPEs with a first first out" buffer memory, which holds in OPEs half of the assigned group of OPEs 103 to 106 and are. by means of a second coupling device 116 to 118, 120,

If any of these buffers (hereinafter referred to as 130 with the other half of the assigned group of FIFOs) is filled, it is assumed that 107 to 110 are connected from OPEs. The first coupling device - for some reason the OCU in question has failed and the device 111 to 114, 119, 129 is in turn no longer able to process the input signals with the second coupling. device of a second OCU N -1 and the second Then, in the same manner as described, the Versa coupling device 115 to 118, 120, 130 is displayed with the first gene and is connected in the same way by transmission of the coupling device of a third OCU N + 1. OPEs to the neighboring OCUs treated. As above, each OCU transmits a display of the monitor device 6 to determine the malfunctions in order to determine the malfunctions, which produce an output signal that records its two failures of the OCU in question. Coupling devices is supplied to a radio

The entire system is constructed in such a way that every assembly fault in the OCU automatically connects the first one at least twice, so that the signals can be redirected to half the group of OPEs 103 to 106 with the second OCU if an error occurs anywhere. As described by N-1 and the connection of the second half of the group of, the OPEs are designed so that their functions 35 to effect OPEs 107 to 110 with the third OCU N + 1 by other OCUs in the event of a failure of the OCU 2 shows the structure of an Nth OCU and its can be taken. The peripheral devices are also connected to eight OPEsl03bisll0 and two adjacent duplicates, so that if one duplicate fails, the OCUs N + 1, N -1 is able to transfer the eight OPEs to two other duplicates, the required functions Groups of four OPEs each to the neighboring OCUs N + 1 continue to execute. Each OCU is programmed to receive 40 Nl in the event of OCU N failure. The transfer of a response via each duplicate of telecommunication cable A, B is analogous to the failure of OCU N + 1, whereby the triple can receive from the units, for which he is to be considered an inquiry group N, N + 1, N + 2, and also sent the same to. The OCU then compares both responses and failures of the OCU N -1, the triplet N, N -1,

makes a decision as to which signal is correct if it is considered N-2.

differentiate. The signals include simple test information, in addition to the components shown in box 100, such as one or more parity digits, to perform a validity includes each OCU power supply and a pulse generation test on the data, so that obviously the signals required by the microprocessor generated, invalid data can be found without difficulty. The eight OPEs 103 to 110 are connected to assigned data. If, on the other hand, an OCU receives responses, the different control units 111 to 118 are connected in the OCU. The latter are on loan, both of which appear to be valid, however, so 50 have certain cache and other functions for which it generates a signal to the operator concerned, for example to check the validity of the operator that they are not processing the data can, because OPEs 103 to 110 received signals and for example they cannot decide which is correct. Among them the execution of certain code conversions. The data conditions the operator must have on the usual control units 111 to 118 are in groups of four each

Use reference books to determine the required information- 55 OPEs connected to common units 119, 120 by means of tion. Any error of this kind will automatically become what the signals from the OPEs are multiplexed. This is transmitted to the monitor device 6; the maintenance person is required to put on a cable line 121 which the main then able to carry out tests to determine what connection between the microprocessor and the section of the system which is closer to it has failed and then re-establish it in connected units . Each unit 119,120 includes operation to set. 60 a FIFO (Fig. 8) for receiving data and command

In addition, the peripheral device 3A, 3B, which also has signals from the OPEs and to enable them to be present twice in time and is connected to the telecommunication cable A, B for processing by the microprocessor in the OCU, has a test unit, which are classified in their inputs.

is switched to ensure as far as possible that the microprocessor itself is in the CPU 122 and the data recorded on the cartridges is correct. In FIG. 65, a FIFO pre-access memory 123 is connected to a) a fast one of the two inputs of the device 3A, 3B by means of the cable line 121, b) a programmable read-only memory is used, and the data stored in both FIFOs is rather 124, c ) a CPU monitor 125 and d) two peripheral ones compared to each other. If the data identity selectors 126, 127 in both FIFOs. There are also options

623 948 6

see for plugging in a printed circuit card a pulse generator, not shown, the pulses of which are fed to (test access) 128 for connecting the fault monitoring monitoring clock generator 131, its unit 7 to the OCU. The CPU monitor 125 is directly connected to the calls stored in the waiting circuit 132 with the monitor device 6 (FIG. 7), and transmits the unit selector interrupt signals. The function of the latter gates 126, 127, each with a duplicate of the telecommunication cable 5, consists in connecting a call A, B after a certain time interval. if necessary to interrupt after warning, for example

Reserve units 129, 130 are connected to one of the two instructions in units of 119, 120 after three prepaid minutes of connection and also to the cable line 121. a public pay phone with coin slot. The unit 129 is connected via a line 102 to the OCU N-1 ben, the monitoring clock 131 comprises a counter, while the unit 120 with the OCU N + 1 ver 10 which is always reset to zero, if the clock is bound. According to this principle, all OCUs that are controlled are connected so that in the event of an interruption in the system. normal sequence of command signals from the CPU 122, after which

Connected to the CPU 122 and, if necessary, into which the clock generator 131 is no longer activated, the counter microprocessor built in (but for reasons of illustration would reach a value which is higher than that normally shown separately from the CPU) is a monitoring clock operation. which provides an indication of the interruption, what 131 and in the quick access memory 123 by means of software then the programs of the CPU 122 are reset to a starting position call interrupt waiting circuit 132 and a peripheral and begin to run again.

Interrogation waiting circuit 133. Connected to the cable line In addition to monitoring the satisfactory radio

121 are maintenance buttons 134 so that a maintenance person of the OCU by means of the clock generator is the monitor device.

can intervene directly in the control of the OCU if this is provided in order to periodically receive signals from the OCU. catch, generated during the program to display for

The CPU 122 is used to process the signals from the satisfactory operation and also statistical data OPEs, making requests to the various peripheral devices regarding the calls that it has processed. Details of the ten can be transferred in the following manner as required. statistical data is transferred from CPU 122 to the buffer in

The operations of the CPU are carried out in sixteen multiplier 125 transmissions (time-plexed time slots assigned to the CPU monitor, of which twelve the maximum twelve slots), so that during the operation of the exchange the status

OPEs are assigned (the eight OPEs that are normally supplied by the statistical data in the monitor device 6-OCU plus four additional OPEs that transmit).

be taken if a neighboring OCU fails). The sixteenth time slot for the maintenance key slot is assigned to the watchdog timer 131 and is available to the 30 134 of the OCU, which allows the maintenance person in the process indicators (indicating that certain time intervals allow failure in the event of failure to the CPU 122 directly have elapsed), two time slots are instructed to the unit selectors and the cause of any errors or out-126.127 assigned, and one is to be diagnosed by the maintenance buttons 134 in the OCU.

that a maintenance person shares data or special, among other storage options for quick access

Command signals can enter the OCU. It can be seen that at 35 memory 123 there is a call memory with 84 addresses, a processor which works in this way, buffer memory one for each call, in which details of the calls have to be used because the actual time for which if given by the operator using or creating any data group from the. When the call is completed, the time slot that the data relates to depends on. If the relevant details from the CPU to peripheral device 3A, signal from an OPE are recorded in one of the buffers in units 119, 40 3B for the load and to monitor device 6 for stati-120, the CPU 122 replies accordingly Transfer purposes. A switchboard memory is the time slot on the signal to record the data is further provided to begin the switches at the switchboard for or their processing in the usual way to relate the call box or the number, i.e. using memory registers in the quick access In the event of a failure of the Nth OCU (either handles memory 123 assigned to the relevant time slot 45 is recorded by the monitoring clock 131 or by the full-to store the data, and initial and intermediate returns of one or more of the input The CPU monitor 125 transmits FIFOs in the data results of calculations as a function of instructions which control units 111 to 118) have derived from the read-only memory 124. A typical signal to the reserve units 129, 130 causing them to be signal from an OPE would be a request for data from the output signals of the units 119, 120 to the reserve in one of the peripheral devices. The CPU would redirect to one in the neighboring OCUs. The CPUs in the ches signal respond by attempting a request OCUs are programmed to forward the reserve units in for the data to the device in question. Check each multiplex cycle to determine whether signals would appear to have passed four additional OPEs to the OCU at that moment. If that's the most accessible, and therefore the query would end up being the case, the CPU addresses the additional four OPEs in the

Queue placed, which is stored in query queue 133 the same way as the assigned eight OPEs, so that is for. When the query finally comes to the head of the queue, the four additional OPEs required additional work, the details from the CPU 122 are fitted into the workflow of the OCU. To this

Quick access memory 123 is selected and the OPE transfer takes place in the standard manner.

Editors 126, 127 assigned to the relevant peripheral device. FIG. 8 shows further details. Two groups with each leads. The data received from the latter would be input by 60 four OPEs via two FIFOs 701, 702 with memory buffers, the CPU 122 into the fast access memory 123, 703 and 704, respectively. The latter each have two outputs from which they can be returned to the OPE. Each OCU that one with the CPU 700 that is connected to the relevant OCU can process 84 calls at the same time, i.e. seven calls are heard, connected and the other of which is connected to each OPE and the reserve CPU stored in the polling waiting circuit 133 belonging to an adjacent OCU, data is in the form of a call box, i.e. a number between 65 is connected. A CPU monitor 705 includes an OR gate 706 see 1 and 84 through which the call in question is identified for the OCU with inputs for the FIFOs 701,702 if it has been filled. with an input for a quarter-second sequencer

The timing of the operations is controlled by 707 in the CPU monitor 705 and with an input for a pulse

7

623 948

generator 708 that provides excitation pulses to CPU 700 to indicate that a pulse or sequence of pulses has failed. Another input 709 is provided so that a maintenance person can manually check the switchover operation. The output of gate 706 is connected to buffers 703,704 5 to control the switching of these buffers to and from the reserve CPUs.

FIG. 3 shows further details of the fault monitoring unit 7 (FIG. 1), which has an active part in the form of a fault monitoring unit (below the dash-dotted line) 10 with a cable line 200 and a passive part in the form of an intervention reserve supply device (above the dash-dotted line) with one Has cable line 220. The components of the passive part serve as an "intervention reserve", i.e. they are repeatedly subject to a sequence 13 of tests under the control of the CPUs 201, 212 during operation of the system, but without interfering with the operation of the system. A maintenance person can therefore remove one of the repeatedly checked components at any time to replace a defective component of the same type, being sure that the selected component is fully operational, while a component in stock could have a fault.

A CPU 201, a fast access memory 202, two read-only memories 203A, 203B and a CPU monitor 204 form a data processing unit, as is used in many units of the system for controlling the tests, which are carried out by the error monitoring unit 7. These components are all connected to the cable line 200. Also connected to the latter are a V24 unit 208, an inter-zone channel 209, which is connected to the peripheral device 4A, 4B, for the purpose of entering statistical data into this device, and a second V24 unit 210, which directly transmits the active part connects to the monitor device 6 for the purpose of receiving its command signals. A special maintenance line from the monitor device 6 is connected via a monitor 223 to 35 of the cable line 200, and magnetic tape cassette units 225A, 225B are provided so that a maintenance person can quickly enter special commands or data for error monitoring procedures. The blocks 226, 227, 228 designate possible positions for other read-only memories 40 for error monitoring programs.

Another V24 unit within the dash-dotted rectangle 211 is also connected to the cable line 200 and, like the first V24 unit 208, connects the passive part to the active part, namely to the 43 cable line 220 via a V24 unit 218 or via a V24 unit contained in a fault monitoring connection unit 222 (top right in FIG. 3). The fault monitoring connector unit 222 can be disconnected from the cable line 220 and connected to the calf line of another unit in the system, so as to enable the error monitoring connector unit to gain access to the components of the unit concerned, so that error monitoring methods can be carried out .

As mentioned, the passive part contains a number of 55 components, for example in the form of printed circuits, which are continuously checked under the control of a CPU 212, a fast access memory 213 and a read-only memory 214, all of which are connected to the cable line 220. Further components of the active part are a 60 peripheral selector 206, a unit 207 common to the OPEs, a peripheral input / output unit 216, an OPE error monitoring unit 217, two data controllers 224A, 224B, an interzone channel 219, an OPE buffer 221 and a monitor input unit 205, which is connected to a CPU monitor 215 63. So that these units can work together in the way they are designed for service operations (apart from connecting the CPU monitor 215 to the

Monitor input unit 205), peripheral selector 206 is connected to input-output unit 216, OPE error monitoring unit 217 is connected to data controllers 224A, 224B, interzone channel 219 and OPE buffer 221 (which is controlled by CPU monitor 215) to effect the switching of this buffer, as described with reference to buffers 703, 704), each have two outputs which are short-circuited.

Both the passive part and the active part have a power supply unit 230 or 232 and a pulse generator 231 or 233, i.e. similar components so that you can standardize their construction.

Such a fault monitoring unit 7 offers the following options:

1. Storage for a range of detail fault monitoring programs.

2. An intervention reserve formation including one with all basic system cards in the intervention state with an alarm if an error occurs in one of the intervention reserve cards.

3. Maintenance access to control fault monitoring programs and to generate the results of these programs as output signals. This access takes place via the remote printer 6A, the monitor device 6 via a V24 connection between the monitor device 6 and the error monitoring unit 7.

4. A possibility for entering and saving secured data groups, eg. B. tariff data, implementation data and credit card stop list data.

5. A possibility for resynchronizing the two real-time clocks in the peripheral device 4A, 4B intended for the tariff.

The fault monitoring connection unit 8 offers the following options:

1. A possibility of interrupting the programs of an addressed CPU in order to carry out an error monitoring program derived from a read-only memory contained in the unit 8.

2. Execution of a fixed value program for checking a basic combination consisting of CPU, pulse generator and fast access memory of an addressed unit.

3. Access for the addressed CPU in order to be able to read the program from the error monitoring unit.

If an error occurs, the monitor device 6 at least indicates which unit, i.e. OCU, peripheral zone or the like has failed, or at best which card has failed. The fault monitoring connector unit is then plugged into the faulty unit. The interrupter is then put into operation so that the CPU begins reading the read-only memory error monitoring program under manual control on the remote printer 6A. If this is insufficient to identify the error, one of the following cards has failed:

1. The power supply, which triggers an alarm separately and accordingly identifies an error by itself.

2. The pulse generator: There is an alarm at the slowest clock pulse, i.e. failure will usually affect the slowest clock pulse and accordingly indicate the error by an alarm.

3. The fast access memory card: Normally two fast access memory cards are used, so that checking with the other fast access memory card can make a good check whether the fault is in the fast access memory card.

4. The CPU.

From this information, it follows that an error that is not already obvious or that can be cleared up

623948

8th

by trying to run through another fast access memory card, usually present in the CPU. The test can be repeated by using the reserve CPU from the error monitoring unit. Another fault may have been caused by the pulse generator, otherwise it would be a very unusual fault in the form of a permanent cable fault in the collecting cable. This can be checked by placing the suspect CPU in the ready position to determine if it is working satisfactorily. If it works correctly, then it is another line fault. Once it has been determined that the CPU, pulse generator, power supply and quick access memory are in order, tests can be carried out to check the read-only memory, the CPU monitor and the input / output devices. In this way, the error can be narrowed down very quickly to a single card. Various cards may need to be removed and replaced if the fault is in the pulse generator or in the power supply and does not operate the alarm devices provided on these cards. Then there may also be an error in connection with the cable line, which may make it necessary to disconnect a number of cards in order to clear up a permanent fault; then you can isolate the faulty card by checking the fault with the cards connected again.

Fig. 4 shows the circuit of a peripheral device except for the magnetic tape cassette recorder 3A, 3B. The peripheral devices differ only in the programming of the read-only memory. As can be seen, the construction of FIG. 4 is similar to that of the other units, such as that of the OCU and the error monitoring unit 7, as far as the data processing part of the circuit is concerned, which consists of a CPU 307, a fast access memory 306, a read-only memory 305 and one CPU monitor 304 exists. However, it is different in that three peripheral input-output control units 301 to 303 are provided and are connected to a cable line 300. Each peripheral input-output control unit is connected to groups of the OCUs.

The circuit of the magnetic tape cassette recorder (FIG. 5) is similar to that according to FIG. 4, but with an additional interzone channel unit 401 and connection units added between the magnetic tape and the CPU and between the magnetic tape cassettes 402 and 403 and a data recorder 404 in the form of a dual tape cassette unit. A unit 415 serves to power the latter.

5, the following units are connected to a cable line 400;

1. The tape cassette unit 404 via the mentioned connection units 403, 402.

2. The interzone channel unit 401.

3. Three peripheral input-output control units 410, 411.412.

4. A CPU monitor 405.

5. A read only memory 408.

6. A quick access memory 407.

7. A CPU 406.

A voltage source 413 and a pulse generator 414 are used to excite all of these units. The interzone channel unit 401 enables the aforementioned comparison of the data arriving at the peripheral device 3A, 3B (FIG. 1) via the duplicate of the telecommunication cable A, B (FIG. 1). A similar unit 401 is located in the duplicate (not shown) of the magnetic tape cassette unit (the peripheral devices are subsequently considered to belong to zone A or to zone B, depending on which duplicate of the telecommunication cable they are connected to). As mentioned, data from one of the duplicates A and B of the telecommunication cable is input to a buffer memory when it arrives at one of the duplicates 3A and 3B of the magnetic tape cassette recording device. They are then compared with the corresponding data in the buffer of the other duplicate 3B or 3A of the peripheral device, and their validity is checked if necessary. These functions are checked by the CPU 406 under program control. Depending on the results of these comparisons and checks, the data is recorded in the data recorder 404.

When the latter is ready for data acquisition, it issues an interrupt signal to the CPU 406. As a result, the latter transfers a data block from the buffer memory to a FIFO, which is located in the connection unit 402. The data recorder 404 then records the data continuously, the reading from the FIFO taking place via the connection unit 403. The format of the recorded data is determined by a control command from CPU 406. The data recorder 404 is of dual design so that the magnetic tape cassette can be replaced without interrupting the recording.

The monitor device 6 of FIG. 6 again uses the same data processing circuit with a CPU 508, a quick access memory 509 and a read-only memory 510 (connected to a cable line 500), but four monitor input units 501 to 504 are also provided here (also connected to the cable line 500), have the direct connection to the rest of the system of FIG. 1 for the purpose of recording the signals from the CPU monitors.

In addition, two dual V24 code connection units 506, 507 are provided in order to connect the monitor device to a remote printer, a punched tape punch and the error monitoring unit. A power supply 512 and a pulse generator 513 are provided.

The monitor device 6 receives data from the OCUs and the peripheral devices, which can be divided into two types. This is statistical and difficulty information from the OCUs and error messages from the OCUs and peripheral devices.

The monitor device 6 generates statistical and difficulty information both printed out and as a paper tape.

Depending on the size of the system, printing is done on printer 6A or on a separate,

not shown printer. If the latter is used, the maintenance person must enter special instructions to do this. The bulk of the statistical data is not regularly required and is only ejected when a mark is set by a maintenance person, via the maintenance printer. The constantly needed information includes:

The total number of statistical messages received by the OCUs for a period of 15 minutes each,

- the percentage of calls answered within 15 seconds (exactly to one decimal place) for fifteen-minute periods (derived from Q counting circles) and

- the average number of vacant positions for fifteen-minute periods.

When an operator starts working on a switchboard, a "put on headphones" signal is transmitted to the OCU, and when the operator leaves the workplace, a "take off headphones" signal is transmitted. The OCU transmits to the monitor device 6 a character string every 16 seconds, which includes the operator stations in operation. The monitor device 6 actuates a pointer for each position upon receipt of this information and averages the number of seats occupied over a fifteen-minute period.

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

9 623948

This statistical information is printed out in a «Statistics printout. In order to make the reference to the two operation buffers »and store them, to simplify all fifteen fields, the maintenance engineer refers to minutes; Every four hours all data in the buffer are printed out with a "device number", which specifies a specific OCU or information about the meaning of the data. Peripheral unit defined and on a «message type».

Part of the monitoring control (effected by the 5 The first 128 types are assigned to the OCUs and the two monitor device 6) is a check of the peripheral 128 of the peripheral devices. It may be appropriate for device 4A, 4B to ensure that the two real-time maintenance engineers have access to the counts and clocks work synchronously. The duplicates of this device 4A, checked in «groups». These groups are defined 4B transmit messages to the monitor device 6 all four through the device to which they belong, e.g. B. Series of OCU hours, including a number in the range 0 to 95, to identify the 10 files related to a particular peripheral unit and the last quarter of an hour in a 24-hour period - the message from that peripheral unit (ie column in the ren and the date. peripheral work file). The message counts can

Just as the information is printed out, will be printed out, deleted, and a threshold will also be punched on a paper strip. A stripe punch can be placed on a group basis that is the same as provided on each installation, but an associated 15 will be more like a column on the row.

Printers are only available for printing in larger systems. To see another control by the maintenance engineer. A conflict can arise between the introduction of the thresholds, the possibility of the function and the statistical function of the monitor being provided in a time-based threshold control, which is direction 6, and this also applies to the fault monitoring unit which enables the maintenance engineer to use a certain 7; a subroutine could be provided to specify a certain threshold count at 20 time periods. In certain cases, the maintenance engineer will do it for you

Difficulty messages are received by the OCU, consider it necessary to set specific maladministration thresholds when the operator has a certain difficulty (eg, regarding a particular message record from either himself or a subscriber) in building a particular OCU. Desires to maintain this threshold setting of a one-phone connection. This informative count is provided in a limited way by Ver-tion is useful for maintenance personnel to identify errors in the telephoto of a separate list of the main fields of work, network. A message is received by the OCU for each when a threshold is reached, the total record count is received and this is ready to be printed for the row or column to see if any other records are saved by the monitor 6. Cyclic drawings are received from the OCU or the peripheral buffer of 4K bytes is used. The printout takes place 30 rie (column check) or the same message type from when the buffer to 3A is filled, namely a statistical OCUs or peripheral devices (column check), printout or on request. In order to support the interpretation if the total number of columns / rows by the maintenance person supports twice this data, the corresponding error record is generated for each difficulty call, the time with respect to the next load.

quarter of an hour, the time being given 35 reports relating to satisfactory operation, when it was received via the monitor device 6 and received by every OCU and peripheral unit every 4 seconds. A possible conflict can be in the same way as that. An 8 second expiry time is set by which the statistical information is treated. grams and if the expiry time has expired before the

Fault records are generated by OCUs and peripheral next report on satisfactory operation of receive devices when a malfunction occurs. The monitor 40 has been, the corresponding message counting device 6 is designed such that it is switched a maximum of 128 by one. A threshold is set in exactly the same message types from OCU and 128 types from the peripheral way as the usual maintenance reports, whose devices can identify errors. received by OCUs or peripheral devices. The error messages are used to advance in a counter field when an error report is generated, the corresponding 8 bits. Two errors are used, a row or column marked to indicate that one such one has been generated for OCU identity versus message type and the other rather report. This ensures that if for the peripheral identity against the message type. Each one further count in the row that reaches the threshold, no row and column has assigned threshold counts, which is then generated by a further report.

Define at which count an error message is to be recorded - Error reports are recorded in a cyclical manner and, to the knowledge of the maintenance person, 50 buffers in which the rows or number of columns and the corresponding should be brought. Usually the threshold counting field (OCU or periphery) are recorded. The values would be fixed, i.e. Part of the program. To ensure the flexibility buffer is large enough to ensure that in general operation, however, the maintenance engineer is able to carry out no transfer. If the sequence is full, set variable threshold levels if necessary. Accordingly, the row or column is not marked to indicate that two sets of thresholds are used per field, 55 a report has been generated, so that the next time the variable threshold is not in use and the fixed receipt of a report , the threshold is again applied as exceeded thresholds. In order to regularly detect interference signals and make another attempt to clear them up, the two fields of error report are entered in the cyclic buffer at midnight etc. Each one is deleted and the counts are transferred to the “historical file” report. The report is then assigned an assigned reference -The maintenance engineer can use the remote printer to enter action 60 number. This enables easier reference control, which can be triggered when the historical file is ordered by the maintenance engineer. When a new error report is updated at night. By inserting the sequence, an alarm is switched off. Dependent

Instruction does not simply overwrite the history of which row or column is affected, it becomes an actual file, but adds up from the last 24-hour period which generates an immediate or a delayed alarm. If

If the maintenance engineer calls up the error report list in the historical file, the counts are switched so that a set of counts can be switched off via the alarms.

can accumulate a number of days. The maintenance device - every quarter of an hour, the monitor device 6 engineer can receive the existing historical files at any time, reports from both tariff peripheral devices, under specifics

623948 10

tion of the quarter of an hour (number in the range 0-95) and that before the operator the connected circuit

Date (day, month, year). If any faulty re-releases. If the called number answers then

is received (parity check) or the two start counting automatically. At the end of the connection,

If reports are more than one minute apart, a fault is generated. If both participants release the line, the learning report is generated automatically in the same way as for the threshold 5 connections and the meter readings. Call data required to connect to the connection

Some details of connecting the CPU monitors 125 to the account are made to the tape cassette (Fig. 2), 204 (Fig. 3), 304 (Fig. 4) and 405 (Fig. 5) with the monitor - Sette transmit, so that the connection circuit is free, gate device 6 are shown in Fig. 7. The output signals to accept a new call.

CPU monitors No. 1, No. 2, No. 6 (for a better overview 10 The tariff period, the fee level and the origin lines are only shown) are stored in assigned FIFO service classes for the service in separate fields 601, 602, 606 which are indicated by scanning signals from that of the VDU, however, the call and monitor input unit 611 will be read out. The monitor destination numbers in a common, assigned input unit 611, 612, 616 each comprise a CPU field. The control of the assigned field is connection 611 A, 612A, 616A assigned to the number of keys from data processing 15, and a flag will be provided in the monitor device 6 to provide data. the first free space of this field is displayed to indicate, The mode of operation of the CPUs in the peripheral devices is which key the control takes over. If no keys similar to that of the CPUs in an OCU are operated as referred to above, the associated field shows the determination described in FIG. 2 by numbering in number if one has been recorded. The time slots are subdivided, each of the inputs or outputs can be assigned an unlimited number of processing transitions which request or supply data. At take. Typical examples are toll-free calls, credit peripheral devices include the input-output card calls, charge transfer calls, personal calls, channels the input-output units, monitoring calls with feedback of duration and charge, and calls with clock and the maintenance keyboard. The operations of entering third-party fee charges.

Magnetic tape cassette recorders also include the 25 A significant saving in labor time for transferring data to the data recorders and that the operator is expected because the mean delay check of the data to be recorded in any duplication between actuation of the key and display by reference on the data that is recorded in the other duplicate of data on the VDU for only a fraction of a second. Because the data processing speed is. Call details, once recorded, that magnetic recording is relatively low (used for both switching and matching with the operating speed of the CPU) are responsible for recording the call and can be used to carry that when the data logger 404 is ready for an unlimited number of repeated attempts to receive data, it will interrupt the CPU's work to establish the connection on each side and cause it to supply selected data. the connection. The charge information, tariff

A typical routine for a telephone exchange rate and charge level is derived automatically from the operator using the described system is the origin and destination code. The way through the following: The manner in which a call is achieved from the circuit to the desired determination, where the medium is received, remains unchanged, and the operator required, answered by an automatic translation of the operator the call as usual. For speaking node office codes.

Acceptance of the call is indicated by the operator 40 although the operator shows the overall control of the “ordinary” service class; if the call has connection time, but in the majority of cases there is a coin machine, the operator can do this automatically. For most calls, if both participants have switched the class to «payphone» by hanging up again, the system automatically carries out the

Actuation of a corresponding «coin operated telephone» button account management function and gives the connection line. When a «change of destination / number» is activated, 45 lines are free so that the lines can be seized for further calls. The VDU automatically shows the applicable ones in the OPE. If usually the operator is on-call tariff period. The “destination number” is now used to record all call details and to be keyed in to the typical national form and has been switched through to the VDU, it no longer needs to be shown. Intervene in the connection by pressing a «Destination entry» button.

a «time stop» button is automatically pressed and the required data processing system is of course suitable. The number is entered into the network. The broker for other purposes, for example for the seat reservation person, then records the exchange of origin / numbering at airlines, at bank service, at Ver-auf, again in the nationally applicable number format and security companies as well as at stock exchange operations. The type of this is also displayed on the VDU. If this recording and number of data handling and / or storage units has been carried out, the call load step is automatically 55th can be chosen arbitrarily, and the same applies to the VDU. As soon as the operator is determined for the data input and evaluation means. It understands that the call has been received in a satisfactory manner, that the number of units and the length of the time is, the "time stop" button is reset and an "automatic that is given is only to be understood as an example" Accounts »button can then be pressed and varied in any suitable way.

8 sheets of drawings

Claims (3)

623948 2 PATENT CLAIMS separate displays of the number of errors of each type to be saved 1. Data processing system with a large number of operators, which relate to each device and each unit (FIG. 6). line stations with data input and evaluation means in 6th data processing system according to claim 5, each station, with a variety of data handling and / characterized in that the monitor device (6) or memory units and having controllers including 5 threshold generating means which provide individual storage means to provide details of a desired data threshold for the various displays and processing responsive to the data input provides an output signal when one of these displays records its gait, wherein the control units exceeds the threshold. Responding to the recorded details selectively 7. data processing system according to claim 5, Query signals and command signals for transmission to the 10 characterized in that the monitor means (6) generate data handling and / or storage units of threshold generating means which threshold a value of the latter in response to the query signals worth the total number of malfunction reports for and command signals Receive data, record the data of each device and each unit and assign an assigned signal to the data evaluation means, produce an output signal if one of the total numbers is recorded by a monitor device (6) which exceeds its threshold value with all 15. Data handling and / or storage units (3A, 3B; 4A, 8th data processing system according to claim 1, 4B; 5A, 5B) and the control devices (10, 20, 30), characterized in that with the monitor device (6), the control devices (10, 20, 30) and the data handling an error monitoring unit (7) with an error monitoring and / or storage units (3A, 3B; 4A, 4B; 5A, 5B) are each connected to a) connection unit (8) which is connected to a central processing device (122, 307, 406), b) a selected data handling and / or memory unit (3A, 3B; programmable read-only memory (124,305,408) which can be connected to at least 4A, 4B; 5A, 5B) or a selected control device (10,20, at least one recurring one, from the central processing device 30) such that defined on the latter or on the latter (122, 307, 406) command program, and an assigned test program by the fault monitor-c) have a monitor unit (125, 215, 304, 405), wherein the control unit (7) can be carried out (Fig. 1). each command program performs a functional test via the central processing device (122, 307, 406) relating to the 9th data processing system according to claim 8, characterized in that the error monitoring unit performs activities and via the periodic triggering of the (7) means (7A) for recording a recording of errors - Transmission of a character string transmitted by the relevant monitor unit (125, monitoring check information in order to include the error monitor-215,304,405), which enables each unit (7) to carry out special checks to carry out the activity of the relevant central processing device 30 (FIG. 1). (122,307,406) specified errors specified, the 10th data processing system according to claim 9, Monitor device (6) the monitor character strings characterized in that the means for recording control devices (10, 20, 30) and the data handling and / is a magnetic tape cartridge. or memory units (3A, 3B; 4A, 4B; 5A, 5B) receives and 11. Data processing system according to claim 1, means for counting the reports of the malfunctions contained in FIG. 35 characterized by a reserve reserve monitor strings advises with a central processing device (212) and separately for each device and unit of the plant. Data processing system associated sub-devices available 2. A data processing system according to claim 1, which is mutually interconnected and program-characterized in that each instruction program is configured to include continuous routine checks to continuously and periodically generate a signal indicating 40 fen, without the operation the data processing system to interrupt the relevant central processing device (122, 307) (FIG. 3). 406) works satisfactorily and that is fed to the monitor device 12. Use of the data processing system according to (6). Claim 1 for recording details of 3. Data processing system according to claim 1, manually established telephone connections, characterized in that each central processing device 45 is characterized in that the data handling and / or storage device (122, 307, 406) is assigned a data memory (123, 306, 407) a main Have data memory (3A, 3B), which is also included, and that it has a monitoring clock (131) by means of one of the control devices (10, 20, 30) for storing individual counters which are at regular intervals a telephone connection, when disconnected, is switched to one by one and leads to a zero value. on a particular instruction from the 50 instruction program, and which also has means responsive to a counter value exceeding a threshold by generate a second signal to which said central processing device (122, 307, 406) responds in order to delete all data from the assigned data memory (123, 306, 407), 55 the invention relates to a system according to the preamble of returning the program to the starting position and patent claims 1 and 1 Use of the facility. to start again (Fig. 2). As is well known, in a telephone network everybody has to do it manually 4. Data processing system according to claim 1, established telephone connection details and characterized in that each control unit (10, 20, 30) is processed in order to correctly charge the telephone account - a group of data input and evaluation means 60 of the subscriber who is subject to the respective fee to charge. (103 to 110), with the data handling and / currently these details being handwritten by the same operator or storage units (3A, 3B; 4A, 4B; 5A, 5B) to all control persons who make the connections councils (10, 20, 30) are accessible (Fig. 2). written on a piece of paper attached to an accounting department 5. Data processing system according to claim 1, characterized for evaluation and for the correct charge, characterized in that the monitor device (6) 65 is forwarded. It is clear that the time it takes to fill out storage means (509) for the malfunction reports of the. the note is necessary, comprises a significant proportion of the data handling and / or storage units (3A, 3B; 4A, operator's working time and consequently the 4B; 5A, 5B) and control units (10, 20, 30) by number of the telephone connections with which the 3rd 623948 Operator can employ. In addition, errors can occur when entering the details by hand, when calculating the fees and / or the booking. Such mistakes often result in a loss of trust in the post office or other body that provides telephone or operator services, and often also in income. It is therefore desirable to have a data processing system that allows all the necessary details to be entered quickly via the telephone connections, and that is extremely reliable at the same time, since it would be unusually expensive to keep the above-mentioned manual filling in addition to the aforementioned data processing system or even duplicate the entire system to train. In addition, it is currently important to check as far as possible whether the latter has received the correct information. The system should therefore have a type of feedback display and also allow any type of test that the information is capable of. According to US Pat. No. 3,898,621, operational safety can be improved, albeit in a very costly manner, by using a duplicate of the main processing device. A monitor device compares the output signals of both duplicates. If any disturbance occurs, the two output signals differ. The monitor device responds to each deviation and then triggers a series of routine checks to determine which of the two duplicates is faulty.