GB2052212A - Call recording arrangements - Google Patents

Abstract

Exchange operator's positions have a keyboard and VDU at which call details are entered by key operations or automatically from monitors, and its own program-controlled processor and buffer store. The operator's positions are connected via data links to common control equipment with suitable interface circuitry at both ends of the data links. The control equipment has its own program-controlled processor and data recording equipment. In use the control equipment, under control of its processor, polls the operators' positions in "multi-drop" manner and polling stops for the call details to be sent to and recorded by the recording equipment. If a polled position has no message to send it replies with a "no- call" message. The absence of a message from a polled operator's position is interpreted as a fault condition.

Description

SPECIFICATION Call recording arrangements This invention relates to call data entry and retrieval equipment for use in, or association with, a manual board serving an automatic telecommunication exchange.

In recent years the incrsased use of processor control in telephone exchanges, plus the spread of subscriber trunk dialling and internationai dialling has reduced the proportion of telephone calls which have to be set up with operator assistance.

However, a substantial number of such calls remains, and indeed will almost certainly always remain. Current practice is that such calls are dealt with by the operator writing out manually an individual ticket for each such call. These tickets are then collected and collated at a central point for billing. Such a system is expensive and time consuming, especially in the environment of an increasingly automated system.

Thus an object of the present invention is to provide equipment for the above purpose which is not subject to the above-mentioned disadvantages.

According to the present invention there is provided automatic call data entry equipment for use in association with a telecommunication exchange, which includes operators' positions from which details of calls to be set up with the help of an operator may be recorded, and common control equipment at which the call details are recorded, in which each said operator's position has a control board for inserting details of a call to be set up, a stored-program controlled processor individual to and at-that operator's position and interface circuitry giving the operators' position access to a data link to the control equipment, in which when details of a call are to be recorded those details are stored under control of the processor at the operator's position in a buffer store at the operator's position, in which the common equipment includes a further stored-program controlled processor, data recording equipment and further interface circuitry giving the control equipment access to the link to the operator's position, in which the control equipment polls the operator's positions under control of the further processor in search of an operator's position whose buffer store contains information relating to a call whose details are to be stored, in which if a polled operator's position has details of a call that position responds by sending those call details to the control equipment, which call details are recorded in the recording equipment under the control of the further processor, in which if a polled operator's position does not have details of a call to be recorded a "no-call" message is sent therefrom to the control equipment, and in which if the control equipment receives no message from a polled operator's position this condition is interpreted as a fault condition.

Such an apparently prodigal provision of processing power is not excessively costly in view of the comparative cheapness of the processors used. These are, in the arrangement to be described, microprocessors in the operators' position and a miniprocessor in the common control position. Thus the equipment is relatively cheap. It has an important advantage, as will be seen later, that if total failure of the common control equipment occurs, the operator's positions can continue to operate for a period.This period ends when the buffer store at an operator's position is full: when this occurs a visual indication is given to the operator, who reverts to manual (ticketing) operation until the common control position is repaired, or retrieves the information from the buffer store and tickets it manually, whereafter the operator's position can be fully used until the buffer store is again full.

An embodiment of the invention will now be described with reference to the accompanying drawings, in which: Fig. 1 is a highly schematic block diagram of an automatic call recording equipment embodying the present invention.

Fig. 2 is a block diagram of an operator's position in equipment as shown in Fig. 1.

Fig. 3 is a block diagram of the software used in an operator's position such as shown in Fig. 2.

Fig. 4 is a block diagram of the duplicated processor arrangements at the control equipment.

Fig. 5 is a somewhat more detailed block diagram of the dualling switch used in Fig. 4.

Fig. 6 is a block diagram of the control position software.

Fig. 1 shows an automatic call recording equipment with a number of operator's positions POS1 to P0596 served by a duplicated common control equipment, in which System A is normally on-line with System B in standby. The operator's positions are connected in batches to a number of links L1 to L32 which extend to the common control equipment. If the common control equipment is remote from the operator's position, these links may include modems, this being indicated by the small crosses. That is, in such case the link serving a number of operator's positions is connected via a modem to the line to the common control equipment, where that link is connected to that position via another modem.

Each operator's position has a dedicated micro' processor which works completely independently of the processor in the common control equipment, so that when the common control equipment is under failure conditions the operator's position can continue to function while the common control equipment is being repaired.

Correspondingly, the failure of an operator's position does not disabie the entire system. The common control equipment also includes a processor, which is a rather larger and more powerful processor than the microprocessors at the operator's position. In the present case the operator's positions use SBC 80/3G microprocessors from Intel while the common control equipment uses a PDP 1 1-34A from Digital Equipment Corporation. The common control equipment acts as a data collection and concentration medium for call records, and as a data base for the storage of information, which is common to a number of operator's positions and is accessible to a number of operator's positions.

Each of the two common control equipment systems includes the above mentioned processor 1. with which is associated a supervisory printer plus keyboard 2. At the "line" side of a system there is a position interface 3 which couples the links L1 to L32 to the common control position, and it will be seen that these links are all connected to both interfaces 3. Further, these interfaces are interconnected via a connection 4, which includes changeover switching arrangements (not shown) which will be referred to later.

Each system also includes a block 5 of multiplexors via which its interface is connected via a single two-way path (known as Unibus) to the processor. This block 5 acts as a multiplexor for one direction of transmission and demultplexor for the other direction of transmission. Also associated with the processor we have a cartridge disc 6, which is a magnetic disc recorder which is usable as a buffer store for call details and/or as a store for the commonly available information referred to above. In addition, there is a magnetic tape unit 7 on which final records are made. Thus unit 7 may be remote from the rest of the system, in which case the connection would include modems, as for the connections between the operator's positions and the common control position.

Each operator's position has a panel with a keyboard for inserting the call details, plus a dial or push-button set for dialling. There is also a video display and associated visual displays, and possibly also alarms. The theoretical maximum of operator's positions which may be serviced by one common control equipment is 480, but in practice this figure is unlikely to be reached as the traffic capacity of the links between the operator's position and the common control equipment would be exceeded.

Thus the the complete system consists of a variable number of operator's positions connected to a common control equipment which includes a support processor, and which is normally duplicated. With two such common control equipments one is normally active while the other is normally in a standby mode. The active support processor serves all of the operator's positions, writing the billing file and providing data/time and data base facilities. If the active processor fails, an alarm is raised and a manual switch is operated to bring the standby into the active role. The system then continues normal operations while the failed equipment is repaired. Other forms of redundancy techniques can be used, e.g. the two systems being in use alternatively for fixed periods, with one processor taking over all work when the other fails.

As already indicated the operator's positions are polled cyclically under control of the processor such as 1 in the active common- control equipment, and when this polling finds an operator's position with a call recording to be sent the polling stops. Each of the operator's positions on a common channel is identified by a unique address number, and during that channel's cycle all positions are polled. Communication is in a half-duplex mode with messages in either direction.

A message from the common control equipment to an operator's position contains that operator's position's number in its header, and this message is received at all operator's positions but is only responded to by the position whose address is in the message header. Such a polling technique is, of course, well known. Transmission of messages from an operator's position is, as will be seen, controlled by the common control equipment. Thus when an operator's position is polled, that position replies with a message containing call recording details in its buffer store, or if the buffer store is vacant with a no-call response. At the control equipment the absence of any respone from a polled operator's position is interpreted as a fault condition, and reported on the control equipment's supervisory printer 2.

All messages in either direction contain character and block parity components which are check on reception of the messages, the checking unit responsing to a message with a positive or negative acknowledgement, depending on the success or failure of the checking operation. In response to a negative acknowledgement, the source of the queried message repeats that message, and this continues until the message has been correctly received and positively acknowledged. Alternatively the system can be arranged to give a fault alarm when a certain preset number of negative acknowledgements have occurred. This can follow established practice in the error detection and correction field.

Note that all call details are stored in the operator's positions and not in the common control equipment, so that failure of the latter does not affect calls in progress. The common control equipment does not receive call details until they are sent to it from the operator's position in question, and on reception these details are immediately transferred to the billing medium the tape unit 7 in the present system. Thus calls in progress can continue to be handled, and new calls set up under operator control, in spite of the failure of the common control equipment. Hence the main effects of an operator's position being unable to communicate with the common control equipment are: (a) the operator's position is unable to send the call details to the common control equipment for transfer to the billing medium, so a queue of such records accumulates in the operator's buffer store.

This may fill the latter, in which case the video display unit shows a message to tell the operator to revert to manual or to empty the store.

(b) the operator's position is unable to effect other transactions, which in the present system includes the processing of booked or deferred calls and certain special facilities. This would be a nuisance but would not effect normal operations which do not need access to the data base.

Failure of an operator's position will, of course, only affect the calls being dealt with at that position.

Fig. 2 is a block diagram showing in essence the arrangement of an operator's position, in which the position controller is a microprocessor which, with its associated circuit elements such as its program memory and the buffer storage is represented by the block 1 0.

The connections to the switchboard include one via an outpulsing unit 11 which is connected to an interface 12, which is itself coupled to the processor block 12, and also to a further interface 13 and a USART 14. This latter is also an interface unit. The interface 1 3 is coupled via a display interface 1 5 to a visual display 16 at the operator's position. Hence signals from the processor or from the common equipment can be displayed to the operator, as can visible indications of the digits and other signals sent from the operator's switchboard.

The USART is coupled via two amplifiers, one for each direction of transmission, to a CPU interface 17: this gives the operator's position access to a link to the support processor at the control position. Such an interface is needed to adapt the information from the USART 14 to the line conditions on the link, and may, as already mentioned, include a modem if the link to the control position is long.

To return to the interface 12, this is also connected to the operator's switchboard via high impedance monitors for the keys or other switches shown as a block 1 8. This connection includes outputs consisting of control signals and a six bit address to select the monitor print to be read and an eight bit output giving the voltage detected at the monitor point. In addition the monitor unit has one clock input 19 derived directly from the processor clock system within block 10.

The process bus is also provided with a multibus interface (24) which is normally used to extend the bus but in this particular case is used to provide additional inputs and output from the processor for reading information from the operator's keyboard (26) and illuminating indicators (27). Note that the monitoring points referred to are telephone subscribers' lines. This connection includes a control connection 1 9 and connection via bidirectional bus-drivers 20. The other connections include the data bus connection 21, between interfaces 12, 13, 14 and the processor, the read/write connection 22 and a select connection 23. The monitors 1 8 are clockpulse controlled, by the system clock pulse source which is used to control the bus connection 21.

The connections 21,22,23 enable the processor to control all operations of the operator's position, many of which have already been referred to in connection with Fig. 1.

We now briefly described the software task structure for an operator's position, and in this respect Fig. 3 will be referred to. Note that as usual with much contemporary software, the software used is modular in structure. Note also the close functional and operation similarities between a software module and its hardware equivalent.

The communications and interrupt task 30 controls the transmission and reception of all messages from the control equipment, and all inter-processor messages within the operator's position. Thus it receives messages from the support processor at the control equipment, indicates an interrupt signal to the processor at the operator's position so that it can handle such messages, and sends back an acknowledgement to the support processor. If a message is correctly received it is deformatted and passed on to be dealt with by the facilities and diagnostic task 31.

The facilities and diagnostics task 31 has two aspects: normall call handling and diagnostics programs. The normal call handling program caters for all the traffic on the operator's position, providing facilities for ordinary calls, urgent calls, mobile calls, etc. as called for by the use of the system. The diagnostic programs provide for a series of simple diagnostic tests to be run on the operator's position hardware, to check that it is functioning correctly.

The next task to be considered is the downline loading task 32, which is the first phase entered when start-up or restart occurs. During this phase the program to be executed during the normal running phase is loaded into the operator's position memory from the support processor. The programs which are downline loaded may be normal running programs or diagnostic programs for use by maintenance engineers to check system operation.

The hardware scanning task 33 scans the switchboard monitor points and the panel keys for inputs, and sends messages to the facilities task 31 whenever an event is detected. This reference to "sends messages" in this context means that the task 31 is then brought into use, which in effect underlines the relations between the hardware and software inplementations of the blocks in Fig. 3. This task 33 also controls the resolution of the monitored signal from a voltage change into a logical signal such as start call, start timing, etc.

The times outputs task 34 receives messages from the facilities task 33 when an output is needed to the portion of the display which gives the operator the "connect circuit" indication, or when digits have to be sent from the operator's position. This task controls the timing for the digit outputs and the periodicity of any flashing signals which have to be given to the operator.

Finaily we consider the buffer store task 35, which is an extension of the facilities and diagnostic task 31. It is used to control the storage of completed call records during periods of failure of the common control equipment or of failure of the link to that position.

facilities task 31 has a store associated with each connect circuit on the switchboard, in which information on the call is stored during the setting up and the progress of the call. At the end of the call, or when the call account key is operated, the connect circuit store is cleared and the information in it is transferred to the buffer store.

Under normal circumstances, the communication task 30 immediately extracts information from the buffer store in response to a polling signal from the support processor at the common control equipment.

If no polling signal is received, the information remains in the buffer store until it is full, which condition is signalled to the operator, who can then empty the store and prepare manual tickets.

This involves reading out the details of each call from the buffer store to the display and preparing tickets manually: Alternatively the data in the buffer store can be left there, being read out when a polling signal is received from the common control equipment after the fault has been cleared.

We now described the processor arrangement at the control equipment with reference to Fig. 4, which shows the duplicated processor arrangement in rather more detail than in Fig. 1.

The two processors 40, 41 are referred to in Fig. 4 as the active processor and the standby processor.

Each processor is a PDP 11 -34A, which is one of the "mid-range" members of the PDP 11 family of processors made by the Digitial Equipment Corporation. Each;processor Is equipped with 64K words of MOS memory, and has a disc file 42 of 5 megabyte capacity, a magnetic tape store 43 for additional information, and a teleprinter 44 for control by and communication with a maintenance engineer.

The processor is also equipped with one or more multiplexors 45, there being four in the system shown in Fig. 1, which communicate with the links to the operators' positions via line drivers and receivers represented by the block 46. The other processor 41 has similar equipment associated with it.

Each of the multiplexors such as 45 is a multiplexed interface for eight or sixteen asynchronous communication channels which, in the present system operate in ASC 11 eight-level code at 2400 band via a suitable interface into the line drivers/receivers 46, which in turn communicate with the operator's position. Each character received is parity-checked by the multiplexor, which sets a status bit when an error is detected, which status bit causes request of a repeat of the queried data. The receive side of the multiplexor includes storage means in which memory characters are buffered, which avoids the need to interrupt the processor each time a character is received. On the transmit side each channel is double character buffered.

The line drivers/receivers in the block of both support processors are connected to the links connecting the central control position to the operator's position, and the output controls on the standby processor are set low so that the drivers' outputs are in the high impedance state and can neither drive nor load channels. The setting of the output controls on the drivers for both processors is under the control of the change-over switch arrangements represented at 47.

The function of the support processor, either 40 or 41 can be summarised as follows: (a) To poll the processor at the operator's position on a regular basis, and receive messages sent from those processors when polled.

(b) To monitor the response to the poll from each operator's position processor and generate an appropriate output report, e.g. via the printer 44, when an operator's position does not respond.

(c) To monitor transmission errors on each link to the operator's position and generate an appropriate output repeat when a predetermined error threshold is exceeded.

(d) To respond appropriately to each type of message received from an operator's position processor, as described below.

(e) To respond to engineer's commands received via the printer 44 to initiate functions such as placing an operator's position on or offline, downline loading programs, etc.

(f) To respond to supervisor's commands received via the the printer 44 for such functions as opening and closing of files on the call billing peripheral (to which the call records are supplied), listing of erroneous call records, etc.

During normal operation, the active processor stores on its disc file the call data records which it receives from the operators' positions. These records are stored by the Message task, which is part of the support processor software, in a portion of the memory designated the spool file, from which the Billing Task is read which creates the billing file. The processor carries out a validation process on each received call record, and that record is only passed to the billing file if it is validated. If the billing file is in a medium with discrete capacity, supervisory commands can be entered via the printer 44 to stop operations while the medium, e.g. a disc or reel, is changed. If the record is not validated, it is not so recorded, and a request for retransmission is sent to the appropriate operator's position. Supervisory commands are also provided for the creation of a duplicate of a previous "volume" of the billing file in the event of medium corruption, or for the output to the printer of (a) validated records, (b) erroneous records, or (c) all records, from an identified contiguous area of the Spool file.

Whenever an operator's position is brought on-line and at hourly intervals while it remains on line, the support processor sends to that operator's position the date and time of day as registered by the system clock.

In systems with optional facilities, e.g. credit card validation or booked calls, some messages from an operator's position will need addition, deletion or look-up of items in a data base on the disc store.

In the system described herein the control position, which includes the support processor is duplicated: we now describe with reference to Figs. 4 and 5 the arrangements for switching from the active to the standby processor when this becomes necessary. Each processor has a switch 48, 49 called a duelling line switch via which it communicates with the other processor, with its own line drivers and with the system alarm 50.

Communication between the switches 48, 49 and the corresponding processor is via one of the channels of the multiplexor such as 45.

Each processor while switched on and loaded with its operational software package is always in the active state, the standby state, or the failed state.

In the Active state the processor assures that its line drivers 46 are enabled, and issues polls on each link to each operator's position in turn, and its outputs billing records to disc and tape. It also polls its dualling switch 48 with an active poll signal. In addition, it outputs billing records to disc or tape and if deferred call facilities are offered it sends details thereof to the standby processor via a dedicated multiplexor channel.

The Active state may be entered automatically on system start-up, in which case that processor is set to its active state whose dualling switch does not receive an "active" signal from the other dualling switch. The active state can also be entered from the standby state at any time after system start up if the dualiing switch no longer receives the "active" signal from the other processor.

A processor leaves the active state as a result of hardware or software failure to the Failed state, or when the dualling switch receives an "active" signal from the other dualling switch as a result of depression of the change-over switch or when as a result of hard disc or tape errors the processor can no longer function: In the Standby state, a processor assumes that its line drivers are not enabled, and it does not poll the links to the operator's positions. It polls its dualling switch with a standby poll, and if the system offers deferred call facilities it receives deferred call details from the active processor.

The Standby state is entered automatically at system start up of the other processor, and it goes to standby when its dualling switch receives an active signal from the other processor's dualling switch. It also enters the standby from the active state when its dualling switch receives an active signal from the other switch as a result of depression of change-over switch.

The Standby state is left by hardware or software failure to the failed state, or by switching to the Active state when its dualling switch detects that an active signal is no longer being received from the other dualling estate.

A processor enters the Failed state when the processor fails or if a "hard" disc or tape error occurs. When in this state, unless otherwise conditioned by the nature of the failure, the processor does not poll the links or the other dualling switch.

The dualling line switch, one of which is shown in Fig. 5, is an electronic circuit unit having eight ports as shown in Fig. 5. Ports 1 and 2 are respectively from and to the processor - via the multiplexor, and ports 3 and 4 are from and to the other dualling switch with port 4 of each one switch connected to port 3 of the other. If the processor is active the dualling switch presents a non-zero voltage level on its port 4, from the Condition outputs unit 60. In other conditions, including power failure, it presents zero voltage.

Port 5 is a driver enable output from another conditions outputs unit 61, which applies a nonzero voltage to enable its processor's driver circuits or zero voltage to disable them. It also prevents zero voltage when the power fails.

Port 6 is the failure alarm port which is connnected directly to a latching input port of the alarm unit 50, Fig. 4, and presents a non-zero voltage while its dualling switch has power and is being polled by its processor within the prescribed time out period. Presentation of zero volts for 10 ms or more enables, via a time out circuit 62, the alarm input, thus sounding the alarm.

Port 7 is an active lamp output for its processor, on the front panel of the processor. This lamp is lit whenever the processor is active, due to non-zero voltage on the alarm lamp port. This port is logically the same as ports 4 and 8.

Port 8 is the changeover alarm port, and the two such ports are gated, as shown at 51, Fig. 4, to the alarm unit 50. If for 10 ms or more neither of these two ports has a non-zero voltage the input to the unit 50 latches and the alarm sounds.

This occurs during a changeover due to failure of the active processor.

A dualling switch also includes a changeover switch 63, which is a non-locking make-beforebreak relay switch operated by a non-locking push button the front panel of the processor. This switch has its input directly connected to port 3. if the button is not depressed the output of the switch 63 is directly connected to the unit's port 3 so that it is at the same voltage threat When the button is depressed, the output of switch 63 is connected to non-zero voltage which, via condition poll response block 64 causes its own processor to go from active to standby state. This eventually switches the other processor from standby to active, so the other dualling switch applies non-zero voltage to the "home" input port No. 3. At this point the active lamp of the other processor is lit and the button can be released.It may be desirable to apply a 250 ms damper on the release action to prevent premature release.

Depressing or releasing the switch 63 of a standby processor should have no effect if the other processor is active. To preserve the non-zero voltage at the output of switch 63 when the other processor is active, the action on both depression and release of the button should be make-beforebreak, with suitable protection circuits between the two dualling switches.

When a processor is in the active or the standby state it polls its dualling switch every 50 ms., the poll being a sequence of three characters, with different characters for active and standby. In the active case, the dualling switch sets or retains ports 4, 7 and 8 at non-zero, which indicates to port 3 of the other dualling switch, to the active lamp and to the alarm circuit that the processor is active. In the standby case, the dualling switch sets or retains ports 4, 7 and 8 to zero, indicating standby.

In the case of both active and standby poll, the dualling switch outputs on its port 2 either an indication the processor that it should become or remain active, or an indication that is should become or remain standby. Which is sent depends on the latest condition of the switch 63, the latter indicates for non-zero and former for zero. The dualling switch -- block 64 thereof -- also responds to either poll by setting or retaining the output of port 6 at non-zero.

if the dualling switch receives no poll within a preset period, e.g. 100 ms., it assumes that its processor has failed and proceeds as for the standby poll. In addition it sets the output of port 6 to zero to enable the alarm, and does not generate any power on port 2. This no poll condition occurs immediately after start-up, and when polling starts thereafter operations occur as just described.

The dualling switch responds to a non-zero voltage on its port 3 - from port 4 of the other dualling switch - to inhibit the line drivers for its own processors. If port 3 carries a zero voltage the line drivers are enabled.

If the output of the switch 63 is a zero voltage, then the next output on port 2 is that for "active processor", while if it is non-zero the next output on port 2 is "standby processor".

The alarm unit 50, Fig. 4, is common to the two processors 40, 41, and is preferably powered by a secure supply independent of the supplies to the processors. It has four inputs, two from the changeover alarm outputs of the two dualling switches and two from the failure alarm outputs of the dualling switches. The first two of these are ORgated. If for 10 ms. or more the alarm unit "sees" a zero voltage level on both of its first inputs or on either of the third and fourth inputs, it engages its latching alarm circuits, which closes two or more sets of relay outputs depending on system requirements. The alarm circuits then remain energized until reset by the operation of a push button 52.

As indicated above, each processor poll its dualiing switch every 50 ms, and the first poll issued by the processor after switch on is a Standby poll. Whether it has issued an Active or a Standby poll, the processor always reacts to the response it receives from the dualling switch by remaining in or changing to the mode dictated by the response from the dualling switch, and the next poll issued by the processor depends on the last response from the dualling switch. If the processor does not receive from the dualling switch a response to poll within a prescribed time and period, it generates a supervisory report. It continues to operate in its existing state, but does not generate any further repeats of the continued failure to respond. However, when a response to poll is received a further supervisory report is generated.

The control position software is, like that of an operator position, modular in nature, as can be seen from Fig. 6. Each of the blocks will now be discussed separately.

The communication task 70 controls transmission and reception of all messages to and from the operator's position processors.

Separation of communication handling into a separate task allows the main Message Task 71 to process discrete transactions in a simple serial manner. This task 70 performs the following functions: (a) It receives outgoing messages from the message task 71 and queues them on a per link basis.

(b) Transmits outgoing messages as soon as the relevant links become free.

(c) Awaits incoming acknowledgement of a message and if not correctly received in time, retransmits a fixed number of times.

(d) When a link becomes free and there are no outgoing messages queued for it, each operator's position "multidropped" off that link is polled in turn.

(e) Awaits a reply to a poll, and if it is an incoming message, receives that message.

(f) If the message is correctly received it is deformatted and sent to the Message task 74.

(g) Fault reports are sent to the Report task 72 via the message task 71 whenever a correct acknowledgement is not received in a reasonable time after a poll or an outgoing message or an incoming message is not correctly received.

(h) After a fixed number of times, an operator's position is regarded as off-line and a report to this effect is sent to the Report task 72.

A special driver task (not shown in Fig. 6) is used to drive the links to the operator's position, using a multi-point protocol with the control status processor as master station and up to sixteen operator's positions as slave stations. This task is responsible for generating, transmitting, receiving and checking headers, parity bits and check scans for each message.

The Message task 71 receives incoming messages from a single input queue, processes them and then outputs a reply message before looping round in its sequence of instructions to handle the next incoming message. This task performs system initialisation, loading and unloading of panel software, call validation, billing record formatting, real-time stamping, and such optional facilities as credit card validation, booked calls, etc.

Aspects of message processing liable to change for each telephone administration, such as call validation and billing record formatting, are separate sub-routines with well-defined program interfaces so that they can easily be replaced.

Such rarely used functions as initialisation and panel loading are designed as overlays so that feature options, e.g. credit and validation, booked calls, may share the same memory space.

Operations by the Message task 71 which produce outputs other than to the operator's position are passed to other tasks for processing via the send/receive functions so that the Messeage task is not held up for transfer.

Billing records pass from the Message task 71 to the Spool task 73 via the send/receive interface, and the Spool task then writes the record to a circular spool file on disc. Whenever the Billing task 74 requests a particular billing record, the Task 73 retrieves it from disc (if necessary) and sends it to the Billing task. It can also retrieve other billing records on request for back-up and listing purposes.

The Billing task 74 creates the primary billing files or magnetic tape 75, and opens or closes these files in response to operator request.

Whenever the billing file is open, the Task 75 requests the next record from the Spool task 73, and as soon as this is satisfied it is written to tape and the next record called for. Whenever a billing file is closed, a record is kept of the file sequence number, and the first and last record in the file, and a report with that information is sent out. The Billing task 74 is also responsible for creating back-up copies of billing files and for dumping particular sequences of valid or invalid records to tape for subsequent analysis.

The Report task 72 controls the output of reports produced by the other tasks, which reports are received via the send/receive interface in coded form and then expanded into a form suitable for printing.

The Command task 76 controls inputting and translating commands from the printer 44. These commands after translation are passed to the relevant task via the send/receive interface.

Commands are provided for: (a) start up and shut down of the system.

(b) control and display the status of each operator's position.

(c) load programs into the operators' position.

(d) control the generation of fault reports for the operator's position.

(e) control the output of billing records and the change-over of the billing tape.

(f) output selected sequences of valid and or non-valid billing records to tape or printer.

When the system is cut in, it initialises and configures itself according to system data stored on disc 43, but it does not communicate with the operators' positions until started by an operator or by a preset timer. The Standby system can be in this dormant state ready to take over from the Active system as soon as the links are switched over.

Some at least of the operator's positions may be provided with automatic and monitoring equipment to enable the entry of call details automatically.

Claims (7)

1. Automatic call data entry equipment for use in association with a telecommunication exchange, which includes operators' positions from which details of calls to be set up with the help of an operator may be recorded, and a common control equipment at which the call details are recorded, in which each said operator's position has a control board for inserting details of a call to be set up, a stored-program controlled processor individual to and at that operator S position and interface circuitry giving the operator's position access to a data link to the control equipment, in which when details of a call are to be recorded those details are stored under control of the processor at the operator's position in a buffer store at that operator's position, in which the common control equipment includes a further stored-program controlled processor, data recording equipment and further interface circuitry giving the control equipment access to the link to the operator's position, in which the control equipment polls the operator's position under control of the further processor in search of an operator's position whose buffer store contains information relating to a call whose details are to be stored, in which if a polled operator's position has details of a call that position responds by sending those call details to the control equipment, which call details are recorded in the recording equipment under the control of the further processor, in which if a polled operator's position does not have details of a call to be recorded a "no-call" message is sent therefrom to the control equipment, and in which if the control equipment receives no message from a polled operator's position this condition is interpreted as a fault condition.

2. Equipment as claimed in claim 1, and which includes two or more links each of which terminates at the common control equipment and each of which serves a number of said operator's positions.

3. Equipment as claimed in claim 1 or 2, and in which the control equipment includes two similar processors each with its own data recording equipment and further interface circuitry, in which one of these processors with its associated equipment operates in an active mode and the other operates in a standby mode, and in which if a fault develops in the active processor or its associated equipment that processor is switched to the standby state and the other processor is switched to the active state to take over control of the system.

4. Equipment as claimed in claim 1, 2 or 3, and in which if an operator's position becomes isolated from the rest of the equipment as a result of a fault condition, that position can continue to function until its buffer store is full, at which point a visible indication is given to the operator, who then records calls manually.

5. Equipment as claimed in claim 1, 2, 3 or 4, and in which the processors are controlled by software whose structure is modular.

6. Equipment as claimed in claim 1,2, 3, 4 or 5, and in which some at least of the operator's positions are provided with automatic line monitoring equipment from which call details may be entered automatically.

7. Automatic call data entry equipment, for use in association with an automatic telecommunication exchange, substantially as described with reference to the accompanying drawings.