JPH033981B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION FIELD OF INDUSTRIAL APPLICATION The present invention relates to supplementary services provided to telephone network subscribers, and more particularly to the delivery of digitally pre-stored voice announcements to these subscribers. This relates to a device for Prior Art Traditionally, telephone network subscribers have dialed appropriate numbers to listen to different voice announcements (e.g., time signals, etc.).
You can listen to news and weather forecasts. Generally, these announcements are recorded on a continuous magnetic tape and broadcast in continuous repetition to several output units (translators). Subscribers can only hear announcements from the point that is read out at the time they access the service. Typically, the subscriber cannot hear the announcement from the beginning, so the subscriber must remain connected until the announcement resumes, at least to the point where it started. This results in a loss of subscriber time and increases line occupancy. Moreover, new subscribers always prefer new types of services. Or in general, new audio services (medical information, daily reports, etc.) and some of the above services require you to listen from the beginning. A device is needed in which the storage device is no longer magnetic tape. These devices must allow subscribers to easily access different announcements and services. It must also be able to guide subscribers with announcements and/or service selection.
This makes it considerably more difficult to manage, making it suitable for providing computer-based devices. Several devices of this kind have already been proposed based on the storage of digital announcements (here "announcement" means any information transmitted in the audio band, such as music). The first device was designed by R.J. Frank (R.J.
“Mass announcement capability” by Frank et al. and TWAnderson et al.
and “mass announcement subsystem”
(“Mass announcement subsystem”). (Both are The Bell System Technical
Journal, V.60, No.60, July-August.1981,
pages 1049−1107). This device has several SAZ systems.
Store a certain number of announcements (or parts of them) PCM encoded at 64 Kbit/s. Each subsystem has two magnetic disks with a capacity of 80 megabytes, divided into sectors corresponding to 30 seconds of audio. The two discs contain the same announcement and are read out 15 seconds apart. Therefore, the same announcement is sent out on the bus every 15 seconds. The output circuitry is synchronized with the announcement repetition rate. This device has several drawbacks, mainly due to the synchronous operation of the system. − It is not possible to memorize very short announcements due to announcements with a duration of 30 seconds. It must be repeated many times until a certain period is reached.
Therefore, it is not possible to store a guidance announcement that directs the subscriber to another service. This feature is desirable when a wide range of services is available. - Due to synchronous operation, the number of announcements that can be stored in each subsystem is limited (specifically 60 announcements). Without restrictions, the average subscriber latency, which is already extremely high (7.5 seconds), becomes even more unacceptable. Similarly, introducing new announcements without increasing latency would require a change in disk read speed, and thus a change in the system. Another drawback concerns reliability,
Depends on the configuration of the device. For example, in the case of a subsystem failure, the subscriber can no longer access the announcements stored therein. Finally, this known system is designed to be connected to an electronic exchange and can only supply announcements to subscribers connected to such an exchange. A second example is described in a paper by DSCheesemen and MBCooper at the International Zurich Seminar on Digital Communications.
“Prospects for voice signaling in the
telephone network”，International Zurich
Seminar on Digital Communications, Zurich,
Parch9−11, 1982.pp.121to127). In the device described in this paper,
32Kbit/s delta modulation or 48Kbit/s
With PCM modulation, the encoded announcements to be sent to subscribers are stored in a solid-state storage device with a capacity of 4 Mbit. Each announcement consists of an average of 10 words, and there are approximately 100 announcements with a total of 200 words. Storage device is 32Mbit/
s is connected to the bus. There are 128 channels connected to the bus, and each announcement is given at a frequency of 250ms. Also, this is a synchronizer. This second device solves the problem of too long average waiting times at the expense of announcement length. In fact, it is designed to provide the subscriber with voice guidance of access to some services, but cannot provide "informational" announcements unless they are very short. Furthermore, the question of service to subscribers connected to electromagnetic exchanges is mentioned but not answered. OBJECTS AND ARRANGEMENTS OF THE INVENTION Contrary to these known devices, the present invention is capable of sending out both very long duration information announcements and guidance announcements for service access, with very limited latency and extremely low latency. It is also possible to provide a very large number of announcements relating to various services, providing a fully asynchronous system with no problems for the introduction of new services and/or announcements. Additionally, splitting the output lines rather than announcements between several units makes service access difficult (or results in call loss) in the event of a failure of one unit;
The above reliability problem is resolved by ensuring that the entire announcement group is never rendered inaudible. Finally, access to the equipment can also be provided by subscribers connected to the electromagnetic exchange. The essential features of the invention are set out in claim 1. Optional features are disclosed in claims 2 to 13. DESCRIPTION OF THE EMBODIMENTS As shown in FIG. 1, a device according to the invention for transmitting digitally pre-stored voice announcements to subscribers of a telephone network RT consists of several similar units U1, U2...Un. (hereinafter referred to as the audio unit). Each unit is connected to the network RT via a certain number of two-way lines.
This line can be used simultaneously with other lines and independently of other lines. These lines are collectively designated E1, E2, . . . En. The number of voice units depends on the number of requests to be serviced simultaneously. Each audio unit uses computers EL1, EL2...
ELn and announcement database AA1, AA
2...Centralized part consisting of AAn and interfaces IEC1, IEC2 towards the line network...
Contains IECn. All possible announcements to be provided to the subscriber are e.g. CVSD (Continuously Variable Slope Delta Modulation).
Data base AA1 for each unit in compressed digital format with Deltamodulation) coding...
Remembered by AAn. These announcements include
Included are informational announcements regarding different voice services or subscriber actions, guidance announcements to facilitate the subscriber's service access operations, and error announcements to inform the subscriber of operational errors. The data base also stores special signals to be used for diagnostic purposes. In these databases, announcements regarding a given service are further organized into comprehensive groups to form a hierarchical or tree structure. The roots are represented by services. The starting point of the main branch is the main service subject. The next branch level section is a more detailed topic. A branch's "votes" are represented by individual announcements. After accessing the service, each subsequent branch level can be reached by dialing numbers. For example, in a health announcement service, the following sequence could arrive at a "child vaccination" announcement. 1. Services of health announcements; 2. Children; 3. Children's health; 4. Disease prevention; 5. Vaccination; bus 1-1, 1- of each computer EL1, EL2...ELn via the normal control interface CD1, CD2...CDn.
2...Connected to 1-n. Audio unit computers EL1, EL2...
ELn handles the delivery of announcements to subscribers so that each subscriber who accesses the service hears the announcement from the beginning and receives the announcement with very low latency, regardless of the length of the requested announcement. must be done. The audio unit computer must also update the data base and monitor the system by detecting malfunctions and disabling faulty components if necessary. To perform these tasks, each unit's computer program consists of four parts or "modules": - an analyzer module that guides the subscriber through the service access procedure, - an executor module that handles the data-based announcement research and transmission to the subscriber, - to which the procedure information is sent at the end of the service access procedure. – a diagnostic module for monitoring the equipment; Access gates IPA1, IPA2...IPAn for announcement updates from remote centers, as well as console CNS1 for human-machine interaction,
CNS2,...CNSn also have buses 1-1, 1-2,...1-
connected to n. The consoles are connected to the bus via serial interfaces IS1, IS2...ISn. These are conventional devices and do not require detailed explanation. Interface part IEC1 towards the line network,
IEC2...What IECn has to do is to store for a certain period of time the announcements or parts of announcements that the computer EL brings from the database, so that they can be heard by the subscribers (in the case of an analog voice switching network). The functions of the system are to reconstitute announcements into formats, to handle telephone connections between subscribers and equipment, and to carry out diagnostic controls for good system operation. The block IEC configuration is shown in FIG.
The unit to which IEC belongs is 32 lines 6(1)...6(8), 6(9)... for connecting to line network RT.
Serving 6(16), 6(17)...6(32). These 32 lines are divided into four groups of 8 lines, each group containing buffer memory M.
1, M2, M3, and M4. Buffer memories M1, M2, M3, and M4 provide an interface between bus 1 of the computer and eight synchronous digital lines. Two such lines 3(1), 3(8) and 3(9), 3(16) for memories M1, M2
It is shown. What the memories M1...M4 must do is the access time and operation time of the announcement data base AA (Figure 1), data transfer from the database to the memories M1...M4 and from there to line 3. By compensating for the different speeds and modalities of the signals, the purpose is to maintain on the line 3 the time relationships on which the audio signal quality strictly depends. Memories M1...M4 are two-port memories. That is, memory that is accessed from a time base within the CO, IE unit computer or memory. Each memory M1...M4 consists of one or more units containing multiple memory elements, each memory element being read and written at an address provided by a computer (asynchronous operation) or a time base (synchronous operation). Can be done. The time base sets the time for two types of operations and sets the time for synchronous operations to send calls or announcements or messages to and from different lines. Each line is a bidirectional line and is unique in that it corresponds to a memory element. The memory element is capable of storing both announcements or messages coming from these lines and destined for the bus, and announcements or messages to be transferred to these lines. The transfer direction is set by the control logic corresponding to each memory element according to the logic value of the first signal supplied from the computer. For each circuit served by a memory unit, the time base is set at the beginning and end of the synchronous operation.
Subject to the computer for suspending and resuming synchronous operation by means of second and third signals received via control logic coupled to the memory elements servicing the line and necessary for synchronous reading or writing of the entire memory element. a first synchronization signal having a period equal to a predetermined portion of time and setting the point of initiation or resumption of synchronized operation associated with the line, and controlling the alternation between the two types of operation of the same portion of each memory element; and generates a second synchronization signal made available to the computer via control logic. Generally, in systems that send out announcements to subscribers, the database is updated by the control center and cannot be updated by the subscribers. Therefore, line 3 should not be bidirectional, and memories M1...M4 can only be written asynchronously and read synchronously. As mentioned above, the announcement database is comprised of ordinary magnetic disks with access times on the order of tens of milliseconds. In order to send a voice announcement of a fixed duration to subscribers, the announcement is divided into elementary sections of a predetermined duration (for example 2 seconds), and the elementary sections are transferred one by one to buffer memories M1...M4. Ru.
The buffer memory consists of two half-memories that operate alternately. If each half-buffer is updated while the other half-buffer is being read, audio continuity is guaranteed. Each buffer memory capacity is 2 seconds and the disk access time is
If it is 50ms, the data transfer time can be ignored compared to the disk access time, so 40 lines can theoretically be serviced. Because the magnetic disk is a random access device, each subscriber line has access to the desired announcement from the beginning and can proceed with listening independently of the status of other lines. Nevertheless, at the start of the announcement, no two seconds of waiting time is required to get the announcement. Disk access requests corresponding to the beginning of an announcement can always be answered within 2 seconds with priority. Although this function is not particularly advantageous in the case of information announcements, it is essential in the case of guidance announcements. Since the guidance announcement involves a large number of subscriber procedures, a waiting time of 2 seconds increases the procedure length and makes the subscriber uncomfortable. In addition to the audio line 3, the outputs of the memories M1...M4 also include a line for providing a clock signal for the transmission rate of line 3 and a line for carrying an availability signal indicating correct operation of the memory and computer. These two lines are collectively represented by connection line 2 (1) for memory M1 and connection line 2 (2) for memory M2. For each memory, the availability signal can be obtained as the logical AND of a signal that makes the memory time base dependent on the computer and a signal issued by a guard circuit that is part of the computer. In such conditions, the availability signal is not present when the computer disables the memory due to a memory failure or in the case of a computer failure. A peripheral unit PA1 is connected between memories M1, M2 and lines 6(1)...6(16), and a similar unit PA2 connects M3, M4 to lines 6(17)...6.
Connect to (32). Only PA1 is shown in detail. Within PA1, lines 3(1)...3(8), 2
(1) and 3(9)...3(16), 2(2) are terminated with respective buffer amplifiers AM1, AM2. The circuit AM is necessary because the lines connecting the memories M1...M4 to the units PA1, PA2 can be very long and highly attenuated and must be compensated for before sending the signals to downstream units. The amplified signals are transmitted to lines 30(1)...30(8),
20(1), 30(9)...30(16), 20(2)
sent to. The audio lines 30 are each line
Connected to units L1...L16. The line unit performs the following operations. - digital-to-analog conversion of the voice present on the line 30 for transmission on the analog line 6 and connection to the line network RT; - extraction from the line 6 of the telephone signal (including dialed digits) from the line network or the line network; - Insertion of telephone signals (especially billing) towards the unit, - diagnosis of the correct operation of the unit's own circuits and the connecting lines towards the rest of the announcement equipment. For these operations, unit L is not connected to line 6, but to line 9 from the test telephone reference generator in order to perform the test on a unit by unit basis. Signal line 20 coming out from AM1 and AM2
(1), 20(2) are all line units L1...L8 (L9...
L16) and is also connected to the signal generator GE. The signal generator GE sends out three digital signals on the connection line 5 as follows: − The bit configuration that represents silence in the coding applied to the audio signal. In the case of CVSD coding, silence consists of alternating 0's and 1's, which is the frequency of the clock signal present on the correct line of connection lines 20(1) and 20(2) divided by 2. obtained by. Memory M1,
If only one of M2 is active, silence is obtained from the clock signal to that memory. If both memories work, you can get silence from either clock signal equally well. - the frequency is within the telephone band, e.g. 1kHz
A bit structure that represents an analog signal. This configuration can occur in read-only storage. Hereinafter, this will be referred to as a "1kHz signal." - a bit configuration representing busy tone according to CCITT Recommendation Q35 (i.e. a signal with alternating "on" and "off" periods at the appropriate frequency);
This bit configuration can also occur in a read-only memory programmed to provide the "off" and "on" period durations required by national regulations. The first two signals are line units L1...L
16 for control purposes and are the same as the special signals described above that are stored in the data base. Third
The signal is sent to the subscriber in certain cases, for example to force the subscriber to hang up the telephone so as not to leave the switching unit in use at the end of the service. The line units L1...L16 are connected via a bus 7 to a peripheral preprocessing unit PE. PE is connected to bus 1 of the audio unit computer. The preprocessor PE receives the picked up telephone standards from the line 6 via appropriate sensors, performs certain processing of said standards, and informs the audio unit computer of the processing results. In the other direction, the PE receives commands from such a computer and sends these to the signal driver.
The signal driver is part of L1...L16 and converts such commands into telephone standards and sends them out via line 6 to the network. The PE consists of two units UCSA and ULS connected in cascade, as described for example in European Patent No. 40412. The unit UCSA, which has the same task as the unit of the same name in the above patent, manages the sensor and the driver, receives the signal from the sensor by the nature of the signal at the time set by the same unit, and detects the state change of the signal to be picked up. Immediately communicate this to the second unit, or at the request of the second unit, communicate the status of other signals picked up to the second unit, and according to the information and/or commands supplied by the unit ULS, the individual driver Or send a command to a driver group. The ULS corresponding to the unit UCL of the above patent follows the evolution of the subscriber's announcement request according to the data supplied by the first unit.
Information and commands are provided to the first unit according to the processing to be performed and the messages coming from the voice unit computer. In the most common case, subscribers U1...Ux can access the announcement device by dialing one of two numbers. Specifically, a "local" subscriber dials a number representing a supplementary service (eg, a number starting with the digit 1 in Italy), and a "remote" subscriber dials a regular number. The criteria and call charges to be incurred are generally different. At this time, the PE must perform different processing in the two cases. However, for better utilization of resources each line unit must be accessible in both ways. To thwart incorrect or fraudulent access attempts (dialing a regular number for a local subscriber and a supplementary service number for a remote subscriber), different acquisition criteria can be used to It must be possible to distinguish between local calls and long distance calls. These two access types correspond to different inputs of unit L and are shown in FIG. 2 for line 6(1) and subscribers UTh and UTj. For example, subscriber UTh is a local subscriber and must access unit L1 via the group selection stage SGs of the supplementary services exchange. Each output of the SGs includes a telephone line voice loop (lines a, b) and an acquisition line (line c).
In contrast, UTj is a "remote" subscriber and accesses L1 via the group selection stage SGu of a normal local exchange. The different output audio loops (lines a ₁ , b ₁ ) of the selection stage SGu are connected in parallel to the corresponding output audio loops a, b of the selection stage SGs, forming the audio loop of line 6 (1). . In contrast, acquisition line c ₁ is separated from line c of selector SGs. Therefore, each line 6 includes four lines (two lines of the voice loop and two acquisition lines). The presence of two different acquisition lines allows the two access types to be readily distinguished. Detection of correct access or incorrect access of a subscriber can be obtained with auxiliary criteria. This is 1
In the case of two types of acquisition (e.g. acquisition on line c, i.e. acquisition by a local subscriber), it must be present on one of the two lines of the voice loop. Because there are two different lines of acquisition, special measures must be taken to ensure that the same line unit is not used by local and remote subscribers at the same time. For information on these measures, please contact Line 6.
In the description of the functions of the sensors and drivers connected to the wire. FIG. 3 shows a typical line unit configuration where line 3 (and thus 30) can carry all announcements provided by the equipment. The multiplexer MX has four inputs, each line 30 and three lines 50, 51, 51 of the connecting line 5 carrying the signals generated by the GE of FIG.
52. The multiplexer switches represent the entire driver of unit L. block
It is controlled by a signal present on the two-wire connection 10 coming out of the AT. The two drivers supplying the selection signals for MX will hereinafter be referred to as drivers T0 and T1. The output 11 of the multiplexer MX is connected to a first control circuit CL and to a digital-to-analog converter CNA. Converter CNA also receives the clock signal present on line 4 of connection line 20. Circuit CL must detect the bit configuration representing silence. Depending on whether the configuration is stored in the GE or in the announcement database, the above bit configuration will be one in line 5.
There may be one (eg, line 51) or line 30. In this way, it is possible to check for correct operation both of the digital data channel coming from unit L and of unit L itself. The result of the check is sent by CL via line 12 to a special sensor (hereinafter referred to as the "logic" sensor) of block SE. Block SE represents the entire sensor corresponding to unit L. The converter CNA is configured according to the audio coding used. The above coding uses Harris Corporation parts.
You can use HC55564. This part is described in Harris Application Note 607.
(“Delta Modulation for Voice
Transmission”) and Dee Johns (D.
Jones). After the converter CNA is the wave and amplification system FA
continues. FA limits the band of the signal emitted from the CNA to the telephone band (300-3400Hz), and transmits the wave signal to line 6.
to the level required for transmission. For the wave section, the block FA includes a low-pass waver, preferably an active waver. Valid cutoff is 6
It can be obtained with the following elliptic wave generator. The amplifier section is composed of an operational amplifier whose gain can be preset. The output 13 of block FA is connected to second and third control circuits CA1, CA2 and a transformer. Control circuits CA1 and CA2 must check the correct operation of the analog part of the line unit. CA1 is a tone detector, GE (Fig. 1)
to the line of connection 5, for example line 52, or
Or the announcement database sends out
The presence of a 1kHz signal must be detected.
CA1 sends a logic signal on line 14a to the SE sensor (the "talk" sensor), the level of which indicates whether a detection has taken place or not. In this way, the conductivity of the connections via CNA and FA is checked. CA2 is a comparator that detects silence,
Silence is indicated by the fact that the output level of the FA is below a given threshold. CA2 sends a logic signal representing the result of the comparison over line 14b to another sensor of SE (the "pause" sensor). Spurious oscillations or noise generated in the CNA or FA that may impair the voice quality of the line 6 can be detected by the CA2. Circuits CL, CA1, and CA2 operate continuously.
The PE samples the signals from the ``logic'', ``talk'' and ``pause'' sensors at appropriate times. Transformer TR produces a balanced voice signal in the telephone loop and decouples the DC signal present in the loop from the analog circuitry of unit L which processes the voice and therefore the voice. The loop coming out of TR is represented here as part of line 15, the four lines correspond to the four lines of line 6 (voice loop and acquisition lines c, c <sub>1</sub> ), and the SE
It also connects to the sensor inside and the driver inside AT. These sensors and drivers, and their related functions will be described in detail with reference to FIG. The four lines of line 15 can be connected to the four lines of line 6 or the lines of line 9 via the switch SW. The control signal for SW is control logic
CSW is provided via line 18. This allows the switch to be operated manually by key AS or by programming in PE. This second
In the case of
Sent to CSW. The logic CSWs of all units L are connected together (connections 16) and work together. This results in
Only one switch can be manually operated at a time, and program-controlled switch operation prevents manual operation of other switches and interrupts manual activation testing. Obviously, the PE program is 2
Prevents simultaneous operation of two switches. The figure also shows that the sensor's picked data is sent to the PE via bus 70, which sends information and commands in the reverse direction from the PE to the driver AT. From the perspective of the PE, the sensors and drivers are a matrix, eg, a matrix with eight element rows, so the bus 70 carries eight-bit words. Driver or sensor addressing is sent from the PE via bus 71 and sent to the SE and
This is done by signals decoded in the logic network DIN connected to AT. Buses 70, 71 constitute connection 7 in FIG. For clarity, the registers that temporarily store addresses and data for the driver are not shown in the figure. The configuration of the sensors and drivers in each matrix will be described after explaining the functions of the drivers and sensors coupled to the line 15 with reference to FIG. In FIG. 4, the sensor is represented by a circle and the driver by a rectangle. Line a is coupled to two current sensors S, R <sup>*</sup> . When a current flows in direction line a coming from unit L (ie, forward ground voltage), sensor S becomes active. When current flows through line a in the outgoing direction (forward battery voltage), sensor R <sup>*</sup> becomes active. These two currents correspond to the dial pulse and hangup signal. Line a is also connected to two drivers A and A <sup>*</sup> . A first driver applies earth voltage to the line and a second driver applies battery voltage. The meaning of these voltages will be made clear in the description of device operation. Line b is coupled to two sensors and two drivers. - BS is a voltage sensor and is active when line b represents forward ground voltage. In the example mentioned (with reference to the Italian situation), this voltage represents an additional criterion for distinguishing between acquisition by a local subscriber or by a remote subscriber, which is necessary to detect incorrect accesses. This voltage is present only in the case of acquisition by a local subscriber. - TAX is a current sensor, which becomes active when current enters the line unit. This current represents the affirmation of charging pulse delivery. - B, B <sup>*</sup> are two drivers similar to A, A <sup>*</sup> . Their functionality will be described in relation to device operation. B <sup>*</sup> must send charging pulses. Lines c, <sub>c1</sub> are coupled to two current sensors C, C1 which are active when current enters the unit, ie when a trap is present. Two current sensors cannot be active at the same time. When unit L is available, the driver
Both wires c, c <sub>1</sub> through the contact point of D <sup>*</sup> have a suitable resistance
connected to the same voltage source via rd. connection 20
Operation of D <sup>*</sup> is possible if an availability signal on the correct line (not shown) indicates correct operation of the memory M and the audio unit computer. Unit L is connected to a group selector which, after detecting the availability of downstream devices, can seize this unit only if a suitable current battery voltage is present. When a unit is not available,
Lines c, <sub>c1</sub> are disconnected from the availability battery voltage source. The configuration of FIG. 4 also prevents "double blocking" of L due to capture of both lines c and <sub>c1</sub> . When there is a capture on either line c, c <sub>1</sub> , the driver
The resistance combined with D <sup>*</sup> is SGs or SGu (Fig. 2)
Together with this, a voltage divider is constructed. The voltage applied to line c (or c <sub>1</sub> ) is different from the battery voltage. After that, SGu (SGs) cannot capture the unit L that SGs (SGu) has already captured. The drivers connected to the lines a, b, c, c <sub>1</sub> of the line 15 and the two drivers T0, T1 controlling the multiplexer MX (FIG. 3) are considered by PE to belong to the same row of each matrix. This means that the data sent to the AT via bus 70 contains the desired state for all such drivers. The driver/“monitor” that controls the SW belongs to a different line. As for the sensors, the sensors ``logic'', ``pause'', ``talk'', BS and another sensor (``fuse'') that detects the presence of battery voltage at the input of unit L belong to the same row. All other sensors coupled to line 15 are placed in separate rows. FIG. 5 shows the control logic CSW1 and CSW2 of the switches SW of two units L. The two logic circuits are identical and only one will be described below.
Two logic circuits need to be represented to show how the priority of single switch operations and program control operations is obtained. In the figure, the line 18-1 for sending commands to the switch connected to CSW1 is the output line of OR gate OR1. The two inputs of OR1 are line 17-1
(program controlled operation) or connected to the output Q of the D-type flip-flop D1. Line 17-1 is also connected via inverter IB1 to line 16a of connection 16 connecting all logic circuits CSW. Line 16a is also connected to the reset input of D1. In the idle state (no test in progress) or manual operation, this line will be a logic one. This connection sets the priority for program control operations, as described in the description of the operation of the logic circuit CSW. The clock input of D1 is connected to the output of a reset flip-flop S1, which is enabled by key AS1. The data input of D1 is connected to line 16b, which is connected to the inverter IA
1 to the output Q of D1. In the idle state, a logic 1 is also present on line 16b.
This connection prevents simultaneous operation of several switch SWs. Logic circuit CSW operates as follows. The unit is not under test and the first operating command is CSW.
Related to 1, if it is a manual command, AS
1 sets the output of S1 to 1,
The output of D1 results in a logic 1 on line 16b. This operates the switch via OR1 and line 18-1. When the output of D1 becomes 1, it becomes 1 via IA1.
The signal at 6b becomes 0. Therefore, operating other keys such as AS2 is not effective. The logic 0 present on line 16b is the output of D2 and subsequently on line 18-2.
This is because it is transferred to If there is no action command from the PE, testing continues until interrupted by a new action on AS1 and the circuit is reset. Line 17- during the test initiated by AS1
If an operation command arrives from the PE, such as via IB2, this command is sent to line 1 via IB2.
D1 is reset by setting the signal at 6a to 0. The signal on line 18-1 becomes 0 again, then switch SW disconnects each unit from DP,
The exam will be interrupted. The logic 0 present on line 16a disables all other logic circuits CSW's D flip-flops and henceforth keys AS.
action has no effect. The scheme of FIGS. 2 and 3 applies if all announcements provided by the device can be transmitted on each line 3 and then 30. In the case of announcements that are requested by a very large number of callers and do not need to be heard in the first place (for example, a "time signal" service), these announcements can be made by assigning each service to a line to which all units L are connected. It may be appropriate to send out continuously. A possible configuration for this variant embodiment is shown in FIG. 6 for the case of n output lines belonging to the same peripheral device PA. For simplicity, the memory is included in the audio unit U, and the block PA is represented only by the line unit multiplexers MXa...MXn. For each multiplexer, reference number 1
0a...10n, 11a...11n are 10,
The line corresponding to the line indicated by 11 is shown. The audio unit U has two parts as shown in the figure.
It is divided into Ua and Ub. Ua can send out the announcements contained in the database on several lines 3a...3m in response to subscriber requests. The second unit Ub continuously sends out to each line 3p...3r announcements that do not need to be heard from the beginning and are requested very frequently. The lines 3a...3m are connected, for example, to a switching line network RC controlled by a computer at AA. RC has n output lines 101a...10
There is 1n (n>m). Each output line is connected to one of the multiplexers MXa...MXn. line 3
p...3r conversely all multiplexers MXa...
Connected to other inputs of MXn. With this configuration, as long as there are free lines 3a...3m, all announcements are taken from section Ua of the voice unit and sent via the line network RC and line 101 to the line unit. All lines 3a...
Once 3m is acquired, announcement requests that do not need to be heard from the beginning are served by Ub via lines 3p...3r. Congestion is detected, for example, by the AA's computer. AA's computer can distinguish between different announcements. Since multiplexer MX has other inputs connected to lines 3p...3r, additional bits are required for multiplexer selection. The same number of auxiliary drivers within the AT correspond to these auxiliary bits (Figure 2). The operation of the device will be explained below. The subscribers UT1...UTx (FIG. 1) access the equipment via a fixed number of dialed digits (access code) which are used to route the call to the voice unit U1...Un which can accept the request.
It is assumed that there is no concentration stage between the network RT and the audio units U1...Un that can generate internal blocks. Thus, once access to the voice unit is obtained, the subscriber can be assured of getting the desired announcement (except in the case of a breakdown), independent of the number of subscribers to whom the same service or announcement is provided. If the line between the RT and the voice unit is not available, the subscriber receives a busy tone just like a normal telephone call. In addition to the access code, the first group of digits dialed by the subscriber may include, in the case of several services (e.g. health information, weather forecast, cooking instructions), one of the services provided by the device.
Contains one or more digits necessary to select one service. This or these digits are not processed by the RT but are sent to the interface of the unit providing the service.
Processed by IEC, more specifically by a peripheral preprocessor. The operation of IEC will be explained in more detail below. For simplicity, it is assumed that the subscriber is connected to unit U1 via line 6(1) (second
fiber). As soon as the preprocessor PE detects an access to the device by analyzing the dial digits, it sends a "seize" message to EL1. Therefore, the analyzer program module is enabled by EL1. After this enablement, the device can start interacting with the subscriber. It guides the subscriber in the announcement selection phase, checks the correctness of the operation, and sends an error message to the subscriber in case of incorrect operation. To send out guidance announcements and error messages, the analyzer module is connected to CD1 (first
Announcement database AA1 via Figure)
access. This program module processes the information that arrives as a signal and orders the transmission of audio information. Different lines connected to one voice unit are treated separately and then sent out announcements to a second subscriber, even if the subscriber is using that unit while serving another subscriber. The steps to do so will be restarted from the beginning. Once access to the equipment and services is obtained, the subscriber must dial one or more digits representing an announcement research procedure. These digits and the next digit are processed by the PE in the same manner as the previous digit. 4
Two steps are possible. - Guidance procedures. The device guides the subscriber in a path along the tree by means of voice messages. This is done by submitting the contents of the next branch level and indicating the digit to be dialed to select the branch at that level. − Direct procedure. This is a procedure that can be used by a subscriber if he knows a priori the digits that represent the passage from one branch level to another until the announcement. The subscriber dials the digit without pausing and receives voice guidance. − Mixing procedure. As with the direct procedure, the subscriber
The digits can be dialed one after the other. If there is an excessive delay between two consecutive digits, the device provides the necessary guidance to the subscriber to move to the next branch level. − Numerical procedures. This procedure is similar to the direct procedure since there is no subscriber guidance. However, the announcement is represented by its own number. This number can be dialed directly without scanning the entire tree. According to the steps selected by the subscriber and communicated by the PE to the analyzer module, that module
To know whether to start transmitting the guidance announcement related to the branch level, to wait for the completion of dialing without transmitting the guidance announcement, or to prepare to send out the guidance announcement completed at the branch level. For example, consider the case of guidance procedures. This has already been done on the computer during announcement research.
Requires maximum intervention of EL1 and database AA1. The devices are bus 1, memory M1, amplifier AM1,
and sends out a preamble to the subscriber via the L1 multiplexer MX. This preamble contains, for example, instructions for selecting a service name and a main branch. This message is converted to analog format by CNA and then
It is transmitted on line 6(1). At any branch level, the guidance is repeated until the subscriber dials the digits or for a maximum predetermined amount of time. As soon as the subscriber recognizes it, he can dial the digit at any point in the guidance announcement. And there is no need to wait for the end of the guidance announcement; if for any branch level the subscriber dials a digit that is not foreseen at that level, the computer EL1 commands the sending of an error message to the subscriber, after which the guidance Announcement resumes. After dialing the final digit and before sending out the announcement, the analyzer module allows the subscriber to check the names of subsequent different nodes, or at least the name of the executed announcement, so that the subscriber can check whether he has actually selected the desired announcement. can be repeated. After this iteration, the subscriber must confirm acceptance of the service by dialing the appropriate digits. Considering that charging normally begins at the beginning of sending an announcement, this prevents subscribers from paying for announcements they do not want. If service is not acceptable, the connection can be restored or the procedure can be repeated from the beginning. Conversely, if the service is accepted, the announcement sending procedure begins. The above explanation also applies to the mixing procedure from the point where the need to provide guidance to the subscriber is detected. For numerical or direct procedures, the analyzer
The module will not be enabled to send messages, nor will it be able to issue error announcements or data repetitions thereafter. In the case of a subscriber dialing error, the subscriber receives an unwanted announcement. If the dialed digit does not correspond to the tree branch, the subscriber will not receive service. Under these conditions, forced recovery will occur. If the subscriber does not hang up after the access procedure, the analyzer module orders the intervention of the executer module. Executor modules manage the actual transmission of announcements and inform such modules of the number of announcements required and the identification of the line to be used for transmission of announcements. In addition, the analyzer module provides the statistics module of the program in EL1 with information related to the operations performed up to that point, in particular the type of access procedure, the persistence time of each tree branch, and when dialing subsequent branches. Report the error that occurred.
This information can then be subjected to further processing out of line. Now considering the operations that the executor module manages, it looks for an announcement on AA1, and when the announcement is retrieved, the module begins sending this announcement over bus 1 to buffer memory M1. The presence of this buffer memory, which stores audio blocks of sufficient duration to compensate for the access time to disk AA1, even when all lines served by unit U1 are in use at the same time, allows the digital-to-analog converter to The audio output from the CNA has the required synchronization characteristics and different lines can be made independent, so each subscriber can receive the announcement from the beginning. After listening, the connection must be restored and made available to other subscribers. After the announcement ends, if the subscriber's recovery time exceeds a predetermined value, the MX
By connecting the output 11 of the line 53 to the line 53, a busy tone is sent to the line. As previously mentioned, the subscriber can interrupt the guidance announcements in order to proceed along the tree. In contrast, error announcements sent out in the event of actual misdialing and misinterpretation by the device of the dialed digits cannot be interrupted. Diagnostic operations can be performed by the diagnostic program module on lines not used for announcement transmission. During this test, the line under test is made unavailable to the subscriber by appropriate commands sent by the ULS. The silence and 1kHz signal stored in AA1 will be used as the test announcement. Using the same route as the voice announcement,
Two test signals are sequentially sent to the selected line, such as 6(8). The test signal is line unit L.
8, which communicates whether the two signals are detected to the diagnostic module via UCSA and ULS. The line is disabled via ULS and UCSA. next,
The test is repeated by using the signal generated by the GE. This allows it to be detected whether the abnormality that has occurred is due to a malfunction in the PA circuit, memory M1, or a malfunction in the connecting lines to these blocks. Depending on the test results, the following diagnoses can be made: [Table] Failure [Table] GE failures can be easily detected because tests on many lines served by the same generator yield negative results. It is very unlikely that several multiplexers will fail at the same time. All detected anomalies are communicated to the operator via the printer and communicated to the statistical program module. Another group of tests concerns the peripheral processor, its connection to the computer of the audio unit, and the good operation of the computer. To control the peripheral processors, the computer EL of the audio unit sends control messages to the ULS. ULS must return this within a specified time. Clearly, a failure to return a message indicates a malfunction of the peripheral processor or line. The UCSA and ULS programs perform the internal controls described in the European patent. An abnormal operation of the UCSA generates an abort signal to the ULS, which blocks the operation and prevents the sending of control messages. For ULS, failures may prevent ULS from stopping. Therefore, the ULS internal error signal can disable message transmission to the computer through a simple circuit. Failure of a connection line to a peripheral processor or computer EL requires disabling all lines controlled by that system. This can be obtained, for example, by setting to 0 the availability signal present on the correct line of connection 2 originating from two memories controlled by the same peripheral processor (then in the example described 16 lines separated). In the event of a failure associated with a voice unit computer, all lines serviced by that voice unit are always disabled using the availability signal. Obviously, the announcement database must be updated synchronously. This is because some announcements are kept for a limited time (eg news, day horoscopes, etc.) or to introduce new services. Two solutions are possible. The first solution is to store the announcements on disk out-of-line and then physically enter them into the computer. This procedure requires a service interruption and is suitable when updates involve a large proportion of announcements. This procedure is also used for initial database information. In contrast, the second type of update can be performed without interrupting service. In this case, a memory similar to the memories M1...M4 (FIG. 2) used in the opposite direction can be used for introducing the digitally encoded speech. This memory constitutes the gates IPA1...IPAn in FIG. When commands are sent through the console, EL transfers the contents of its memory to free space on its disk while continuing to send out old announcements. The change between the old and new announcement is made simply by instructing the executor module to pick up the announcement at a different address at a given time when a console command is sent again. The operation of the peripheral units is explained below using the flowchart shown in Figure 7.
This will be explained in detail with reference to the chart. In these flow charts, some operations are shown only once, although in some situations some operations may be repeatable. The connector for each display is represented by a Greek letter followed by a number representing the phase of operation. For simplicity, line 6 is also used.
This will be explained with reference to the announcement sent by unit U1 via (1). When the announcer is placed in service, the preprocessor PE (Figure 2) disables all line units (Phase 0 of Figure 7).
This corresponds to disconnecting lines c, c ₁ (FIG. 4) from the battery voltage and is necessary to allow the operator to configure the device correctly. The line unit is made available via a message generated by the diagnostic module of the voice unit computer program upon operator command (set enable, availability from DIAGN). Availability status is connection line c, c ₁ (Figure 4)
, and line a corresponds to the battery voltage. The last connection made via driver A allows the selection to be detected. As a result of these operations, the preprocessor moves to a free and usable phase (represented by 1 in FIG. 7). This phase corresponds to the idle state of line unit L. An idle state is left when a service access request is detected. The demand is represented by the acquisition criterion of line c or _c1 . The evolution of the requirements will be explained below. All operations shall proceed normally until the end. The abnormality will be explained later. Up to a certain processing phase, both line acquisitions are treated similarly, except for abnormal conditions. Therefore, no distinction will be made below. After an acquisition criterion is detected due to a change in the state of sensor C or C1 (Figure 4) and signals acquisition on line _c1 (setting the long distance call flag, Figure 7), if this is true, the PE will request the acquisition of that program module (analyzer) on the computer. It processes the service request (requests ANS acquisition) and calls the status of the sensor BS (Figure 4) to detect whether the access request is valid ( calls the status of line b, Figure 7). As previously mentioned, a correct access request by a local subscriber requires the dialing of a number representing an auxiliary service, resulting in the capture of line c and the auxiliary criterion of line b detected by the sensor BS. Correct request for a remote subscriber requires dialing the normal subscriber number, resulting in the capture of line _c1 . Dial of a number of an auxiliary service by a remote subscriber (capture of line c in the absence of an auxiliary standard of line b ) or dialing of an ordinary number by a local subscriber (of line c ₁ in the presence of an auxiliary standard of line b ) On the other hand, capture) is considered a fraudulent access attempt. Then the rendezvous phase (2 in Figure 7) is reached. There unit L1 is captured. The response from the EL1 program (availability from the ANS) must arrive within the given time. Forced restoration will be performed according to the format described below. The same situation is repeated in other phases waiting for program intervention in EL1 (Figure 1). While waiting for a response from EL1, SE (Figure 3)
can send out information about the state of the BS (the state of line b ). If this state is correct, a response from EL1 is still awaited, and if it is incorrect, forced recovery is performed. During the meeting, dialed digits (digits from subscribers) may also arrive.
These digits are detected by a sensor S with an active driver A. These digits are stored (store digits) until a response arrives from EL1. After the response by EL1, when those numbers are sufficient to detect the requested service, digits are sent to EL1 via a message for each digit (digit to ANS), after which the number is completed. followed by a character indicating (service code end). Next phase of waiting for a message (Phase 5)
, where the analyzer module indicates the type of procedure required by the subscriber. If a digit arrives from a subscriber during an appointment, the digit is stored. Further processing of the digits depends on the procedure selected. When a message indicating the procedure (characterization from the ANS) arrives, in the case of a numerical procedure, phase 5 is maintained, waiting for the subscriber to complete the dialing. This fact is detected by EL1. EL1 sends a message to the PE indicating the end of the acceptance phase of the service request. As a result, the PE precedes the announcement sending operation. The execution waiting phase (phase 10) is reached. In the case of a direct or mixed procedure, the stored digits, if any, are sent to the ANS in the manner described for phase 2 to reach the management phase of the procedure (phase 7). In the case of a guidance procedure, the digit is ignored (reset the digit area). This is because the subscriber must receive the first guidance announcement. Furthermore, since the management of the two cases is different (phases 9 or 8, respectively),
The PE must check whether the procedure always requires the same number of dial digits (fixed digits) for different steps. In phase 7, if the procedure is direct, the subscriber's dialed digits will be sent as soon as they arrive.
Sent to EL1 (Figure 1). At the end of the dial,
Similar to the numerical procedure, phase 10 is reached. In the case of a mixed procedure, the PE can receive messages in the order of setting up the voice path towards the subscriber, and the subscriber can receive guidance announcements or error announcements. This message is sent when the time between digits exceeds a certain limit. As a result, the PE sends the logic states necessary for connecting the line 11 to the line 3 to the drivers T0, T1 in the AT (FIG. 3) (setting up the audio path). Upon reception of each dial digit, the PE checks whether a message is being sent and whether it is interruptible (eg a guidance message) or uninterruptible (eg an error message) (presence/interruptible voice). In both cases, the digit is EL1
(digit to ANS). In the first case the audio path is interrupted (audio path restoration) and in the second case it is not interrupted. In this second case, the computer rejects the received digit. At the end of dialing and when the acceptance phase end message arrives, if the voice path has been set, phase 10 is reached after the voice path is temporarily restored. This behavior is appropriate. This is because subsequent announcement transmission is managed by different modules of the EL1 program, and a certain amount of time must elapse between the end of the acceptance phase and the moment when the announcement is ready to be transmitted on line 3. (on the order of 10 seconds or more). If the voice path remains established at this time, the subscriber may receive line noise or guidance announcement sequences that are still being transmitted before the analyzer module is actually restored. In the case of guided order with a variable number of digits (phase 8), the PE receives a message setting up the voice path towards the subscriber, as in the mixed procedure. Furthermore, when a digit arrives from a subscriber, EL1
is sent to check whether the audio can be interrupted. After the transmission of each digit, a guard time count is started to distinguish between digits relating to the same parameter and digits relating to different parameters (inter-parameter time counting). The word "parameter" refers to the number to be dialed in response to a guidance message. If the next digit arrives before the time has elapsed, it will indicate the same parameters and then follow the procedure described above. If time elapses before the digit arrives (interparameter time elapsed), the PE sends a parameter end character to the ANS. At the end of the dial, an acceptance phase message arrives and is managed in the same way as in phase 7. Even in the fixed digit guidance procedure (phase 9),
As soon as the digit arrives, it is sent to the ANS. However, in this case, there is no need to count the guard time. This is because the number of digits that make up the parameter is known. A parameter end message is then sent after the number of digits. Phase 10 is a waiting phase for intervention of the executor module of EL1. The executor module first sends out two consecutive messages. The first message (Availability and Billing from TES) indicates the availability of the module and tells how to charge for the service. Charges depend on the type of announcement and/or capture. In detail, this message includes a "type" (fixed registration of response, periodic registration during listening, fixed registration at the end of announcement), "base" (time unit to be used in periodic registration), , e.g. 1 order), "value/time)" (number of pulses to be sent for fixed registrations, or registration rate for periodic registrations). If not, a message indicates this condition. A second message (response from the TES) indicates that the announcement is ready. In this respect, the procedure for local and long-distance calls is different. This is because the registration form and telephone standards to be sent are generally different.The type of call (long distance)
A check is performed only if the subscriber has not recovered. This case will be discussed later. Considering first the local call, the registration has to be performed by the auxiliary service exchange according to the information provided by the PE. If a charge has to be made for an announcement, it is assumed that registration is done at the time of response. The value/time parameter of the registered message includes the number of pulses to be selected. In the case of a free announcement, the multiplexer MX is activated immediately to configure or reconfigure the audio path and reach the next phase (Phase 1).
1. Local call processing). In the opposite case, a first registration pulse is sent out via the driver B ^* (FIG. 3) before the setting or resetting of the audio path. Then, counting of the waiting time is started. Within that time, sensor TAX must pick up a pulse affirmation. In phase 11, a positive registration pulse is waited for. If an affirmation arrives, an appropriate signal is sent (setting the flag OKTAS). When the time has elapsed, another registration pulse is sent in the manner described above, if necessary, so that listening to the announcement can begin. This phase ends with recovery by the subscriber or by the computer at the end of the announcement. For computer recovery. An announce end message is received (stop from TES) and the line unit is made unavailable for a certain period of time (unavailability, count of availability waiting time). This allows all units that have been connected to be restored, reaching the intermediate phase of restoration (Phase 6). When the time has elapsed, the unit is re-enabled and reaches Phase 1 again. That is, the PE can process new requests for that line unit. If the subscriber's recovery precedes the computer's recovery (for example, if the subscriber interrupts listening before finishing), this event is communicated to EL1 (stop to TES) and the announcement sending or processing is stopped. Line unit is disabled and EL
A count of the waiting time for the affirmation of one processing step is started. After that, phase 3 is reached.
If a positive has arrived, the line unit is re-enabled (δ3) after checking that no diagnostic tests were required. If no affirmation arrives, availability is set again after the alarm occurs. For long distance calls, the originating switch sends out registration pulses. For this reason, the calling exchange must undergo end-of-selection and answer criteria. For a given time, PE
During phase 10, the end-of-selection criterion is sent (end-of-selection activation). Thereafter, the audio path is established or re-established while waiting for the criterion to end. (Counting End of Selection Time, Phase 12), at the end of the criterion, drivers A ^* and B are enabled for a certain period of time (Intermediate Phase) and counting of that time is started (Counting Intermediate Phase Duration Time); Wait for time to pass (Phase 13).
When the time has elapsed, the PE disables all drivers on lines a and b that sent the answer criteria and reaches phase 14 in response to long distance call processing. This phase is similar to the phase for local calls, and EL
1 or by the subscriber. In the first case, upon receiving the message, the PE can send out clearback criteria via drivers A ^* and B and begin counting the waiting time for subscriber restoration. If subscriber restoration arrives within the given time, the actions following point δ3 are performed. Otherwise, an alarm will be sent before these operations. If the subscriber is restored before the computer, the same pattern applies as for local calls (γ10). Forced recovery may occur during access or execution of the service. In the above case, ie, no response from EL1's program and an incorrect access attempt, forced recovery may be due to a processing stop requested by EL1. This stop can be requested during both the service acceptance and service execution phases. In the case of a condition requiring forced recovery in the acceptance phase, end-of-selection criteria are created in the same way as described for long-distance calls (phase 2, γ2). At the end, the unit is rendered unusable as in the case of normal recovery. The procedure is shown for phase 2 only. The operation of the end-of-selection criterion for both call types avoids capture-type checks. On the other hand, the end-of-selection criterion for local calls does not affect the operation of the switch. Forced recovery by processing stop during service execution does not require sending an end-of-selection standard. Such a criterion is not necessary for local calls, as it is already provided for long distance calls. In this case, the action follows the action item ω ₄ . Other than cases where forced recovery is required, other abnormal cases are disable by the diagnostic program module of EL1 (disable from DIAGN) or subscriber recovery before termination of service. Diagnostic Program Disable (only shown for Phase 1 for simplicity), the line unit is disabled and Phase 0 is resumed. If disable is commanded in the phase in which the line unit is captured (phase 2 and below), a forced recovery must occur before disabling. Regarding subscriber recovery, a distinction must be made between recovery that occurs during the acceptance or execution phase and recovery that occurs in an intermediate phase between acceptance and execution (phase 10). Recovery of intermediate phases is treated as recovery of execution phases. Recovery is performed only after a response by the executor program module of EL1. When recovery occurs during the acceptance procedure (Phase 2), the PE disables the line unit and sends a message requiring ANS Stop. Next, wait for a stop affirmation (phase 3). The line unit is rendered unusable. Otherwise, a subscriber may seize a line unit when other circuits are not yet free. next,
The procedure proceeds as before. So far, when describing the evolution of a call, the preprocessor has managed the test in terms of the telephone in which the DP occurs;
Diagnostic control is also performed using silence and a 1kHz signal. A message requesting a test (switch from DIAGN) can arrive at any time and is sent by the operator. A distinction must be made between tests for free or unavailable line units and tests for busy line units. The line unit is busy (so
PE is a phase different from phases 0 and 1). The PE replies that the test cannot be performed (response NOK to DIAGN); in fact, the test requested by the program is a routine operation that should not affect regular service. For non-routine cases, tests can be requested manually and can be run even when the unit is busy. If the unit is in phase 0 or 1, after affirmation of feasibility (response to switch OK), the PE
Connect the connecting lines 9 and 15 by enabling SW via the appropriate drivers of AT and CSW (FIG. 2) (test direction=yes, monitor towards test). At the operator's request, all predicted types of normal telephone standards are sent to line 15 by the DP. The PE performs the processing and the operator can check the correctness of the operation with the help of visual signals and tracing. When the test is finished, the preprocessor PE again reaches the start phase for a new message. The same thing occurs for manually ordered tests. Diagnostic controls may also be requested by the operator and may only be performed while the line unit is idle or unavailable. The operator can request that the control be executed only once and indicate whether it must be executed immediately or as soon as possible. Periodic control cycles may also be required. The operator's request is sent via an appropriate message.
Sent from EL1 to PE (preset to test).
In the case of periodic control cycles, a message is sent for each control action. Messages can arrive in any phase. For messages arriving in phase 1 (line unit free), the PE responds to EL1 with a "ready" message, makes the unit unavailable to the subscriber, and moves to phase 15 (Figure 7E). Control is actually executed here with explicit commands from EL1. If the "preset to test" message arrives while the PE is in phase 0, then after generation of the appropriate signal, phase 15 is reached directly (set flag test in phase 0). If the message arrives in a phase other than phase 0 or 1 (for example, phase 2), the PE
checks whether immediate execution is requested. If not, the PE, whose presence is checked after recovery (see phase 3, δ3), generates an appropriate signal (setting the test flag) to inform EL1 that the unit has been preset for test. If immediate execution is requested, PE informs EL1 that control cannot be executed (response to DIAGN).
NOK). In phase 15, after receiving the execution command,
The PE detects whether the "silence" and 1kHz signals sent by U1 are correctly received by reading the states of the "logic", "pause", and "speech" sensors for the first time (reading the test bit state). . When a read result is requested (test bit state), PE communicates this to EL1 (DIAGN
test results). The PE then receives a new execution order (tone activation from DIAGN) and performs line 11 by enabling drivers T0, T1.
First there is silence, then the 1kHz signal stored in the GE is sent out (special tone). The operation proceeds in the same way as for the signal coming from U1. If necessary, intervention is carried out by the diagnostic program module of EL1. Once the diagnostic control is complete, EL1 must send an "end of test" message to the PE. This message causes PE to enter the start phase (phase 0 or 1). Therefore,
Along with the execution request or preset message, the PE must receive information related to the time that EL1 has to use for control execution and interpretation of results. Obviously, this time is calculated within very large tolerance limits. If time elapses before the arrival of the "Last End" message,
PE voluntarily exits Phase 15. After the "Test End" message, if the control fails, the PE
disables the unit. Conversely, if control has a positive result, the unit is re-enabled and the PE moves to phase 1 if control is executed in that phase, and disabled if control is executed in phase 0. It works the same as in the case of . Although the above explanation has been made using non-limiting examples,
Variations and changes may be made without departing from the scope of the invention as defined in the claims. For example, if the telephone line 6 is a PCM line, the entire blocks CNA, FA, and TR in FIG. 3 are replaced with a device for converting CVSD encoded voice into PCM encoded voice. Circuit CA2 may be omitted. The sensor and driver configurations are also modified.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the device. FIG. 2 is a block diagram of the interface with the analog telephone network. FIG. 3 is a block diagram of the line unit. FIG. 4 is a connection diagram of a telephone signal sensor and a driver. FIG. 5 is a circuit diagram of the line unit logic circuit. FIG. 6 is a partial circuit diagram of a modification. FIG. 7 is a flow chart of the preprocessor. Explanation of codes, U1~Un...Audio unit, RT
...Line network, M1 to M4...Memory, PE...
Peripheral preprocessing equipment, PA1, PA2...Peripheral units, L1...L16...Line units.

Claims (1)

[Claims] 1. At least one unit U1, U2...Un
(hereinafter referred to as the "audio unit"), which consists of a plurality of independent lines 6(1)...6
(32) through the telephone network RT without intermediate concentrator stages.
connected to a switching device in the telephone network, through which each line can be assigned to a subscriber UT1...UTx requiring announcements, said audio unit: - all announcements to be provided to the subscriber; announcement data that is stored in digital format.
bases AA1...AAn, the access to which is independent for each subscriber connected to the voice unit; - control interfaces CD1...CDn; A computer EL1 with a bus 1-1...1-n to which the announcement data base is connected
...ELn, a computer EL1, which manages the sending of announcements to subscribers so that subscribers accessing the device receive announcements from the beginning, updates the database, and monitors the device by disabling faulty parts. ELn; - Bus 1-1...1 of computer EL1...ELn
- an interface IEC1...IECn between the circuit network RT connected to n and the audio unit itself, where the computer EL1...ELn
Base AA1...Interface IEC1 that temporarily stores announcements or parts of announcements retrieved from AAn and sent to subscribers.
An apparatus for transmitting digitally pre-stored voice announcements including IECn; to a telephone network, wherein the announcement data base AA1...
AAn stores information announcements and diagnostic control signals in addition to guidance announcements for guiding subscribers regarding interactive access operations, and the computers EL1...ELn manage the sending of announcements and With this, the subscriber who has accessed the device can receive the desired announcement after a short waiting time, which is done as simply as in the case of a two-way conversation procedure, and this is done by the subscriber who receives said announcement. The above-mentioned interfaces IEC1...IECn provide means for managing the telephone connection between the subscriber and the equipment and for diagnostic control regarding the correct operation of the equipment. and the means include one or more peripheral units PA1, PA
2, PA, each peripheral unit containing the following equipment: - a plurality of synchronous digital audio lines 30(1)...
30(16) and at least one corresponding line 6(1)...6(32) for connection to the line network
A plurality of line units L1...L16 are respectively inserted between the connection line 6(1) and the signal addressed to the line network RT.
...6 (32) to extract from the line the signals coming from the line network RT; test and control signals generated in the peripheral units or coming from the announcement database AA1...AAn of each voice unit. A plurality of line units L1...L perform diagnostics on the correct operation of the interface and its connections with devices using
16; - one or several memories M1 on one side...
For control purposes, it is connected to the line 20(1)...20(2) carrying the clock signal and correct operating signals sent from M4, and on the other side connectable to the line unit L1...L16 instead of the digital audio line 30. means for generating a digital signal for use in GE, the digital signal generating means being such that the transmission of the control signal is disabled in the absence of a signal representative of correct operation of the device;
GE; − Line 6(1)…6 to connect to the line network
Test telephone reference generator DP connectable to the line unit instead of (32); - on one side all line units L1...
A peripheral preprocessor PE connected to L16 and on the other side to the computer EL1...ELn of the audio unit, with telephone signal standards supplied from the network RT or test standards issued inside the peripheral unit. , and a signal indicating whether the presence of a control signal is detected on said line.
Line unit L1 sends messages to the computer regarding the information received from the unit and represented by the criteria and/or signals mentioned above.
...receives messages from the computer regarding the operations to be carried out in L16; for connecting the line network and for connecting the same unit to the control signal and test standard generators GE, DP; Apparatus for transmitting digitally pre-stored voice announcements onto a telephone network, characterized in that: 2 comprises several voice units, each voice unit being connected to the telephone network RT independently of each other and independently of the connection lines of the other units, and
2. The apparatus of claim 1, wherein each unit's announcement database contains all announcements to be provided to subscribers. 3 The interface equipment IEC1...IECn connected to the line network RT further comprises: - on one side connected to the computer bus 1 and on the other side a group of synchronous digital lines 3;
(1)...3(8); 3(9)...3(16), and furthermore, a line 2 ( 1), 2
(2) also connected to one or several buffer memories M1...M4, which buffer memories are used to record data bus access times and announcements or parts thereof before being transmitted to the subscriber. Restoring transmission continuity to the synchronous line by storing the announcement or part thereof for a time at least equal to the total time with the readout time, the computer of the unit U1...Un for storing the announcement or part thereof; and memory M for transmitting announcements or parts thereof to the synchronous line.
one or several buffer memories M1...M4 whose memories are accessed by time-based commands within 1...M4; and - amplify the signals issued from the buffer memories to lines 2, 3. means AM1...AM2, a digital audio line 30 and a signal line 20 respectively coupled to the output line of the memory;
Amplifying means AM1 sends the amplified signal to...
An apparatus according to claim 1 or 2, comprising AM2. 4 Digital audio lines 30(1)...30(8), 30(9)...30 that transmit amplified audio signals output from the same buffer memories M1...M4
Line units L1...L8, combined with (16)
L9...L16 are all connected to signal lines 20(1), 20(2) carrying an amplified clock signal and a correct operating signal to the buffer memory.
The apparatus according to any one of Items 1 to 3. 5 Each audio unit U1, U2...Un connects to the access gate IPA1... connected to the computer bus for updating the announcement database.
IPAn, and the access gate IPA1...IPAn is configured from a dual-ported memory that synchronously updates announcements or portions of announcements upon arrival of an internal time-based command. 5. A device according to any one of claims 1 to 4, wherein the device receives and stores and is read out asynchronously upon arrival of a command from a computer EL1...ELn. 6 to each line unit: - a multiplexer MX whose first input is connected to the audio digital line 30 and whose at least second and third inputs are connected to the digital control signal generator GE; a multiplexer MX which receives from said control signal generator a signal representing silence in the coding for the signal and a signal whose frequency is within the telephone band, the switching of said multiplexer being controlled by a preprocessor PE; - a local generator; GE or announcement data
First and second control circuits CL, CA1 connected to the output of the multiplexer MX for detecting the presence of silence coding signals sent by the bases AA1...AAn and signals in the telephone band, the logic representing the detection; Control circuit that sends signals
CL, CA1; - a line for connecting the line unit to the line network RT or the test reference generator DP, on one side to the output of the multiplexer and on the other side to the line network RT and the test telephone reference generator DP; switching circuit SW connected to one of the switching circuits SW; - a logic circuit network that controls the switching circuit SW;
CSW, and the logic circuit network connects the switching circuit SW to a preprocessor PE or a manual control device.
AS: Allows operation by AS1 and AS2, logic circuit network CSW of all units
are connected together so that only one switching circuit SW can be operated manually at a time;
Logic in which the operation of SW prevents the manual operation of other switching circuits and the line units L1...L16 connected to the test criterion generator DP via the manual operation of the switching circuit SW are disconnected from the test criterion generator. network CSW; - connected on one side to the outputs of the control circuits CA1, CL and to the lines of the line for connection to the network to receive the logic signals emitted by said circuits and the telephone standards present on the line; on the other side; - a sensor set SE connected to the preprocessing device PE on the side of the sensor set SE and transmitting data represented by the sensed signals and/or criteria to said preprocessing device; On the side the multiplexer MX, the control logic CSW of the switching circuit SW and the line network RT
A driver set AT connected to the connection line to the preprocessor PE and on the other side to the preprocessor PE, in which the individual drivers are connected to the multiplexer MX or the switching circuit from said preprocessor.
a driver set AT that receives information and commands related to the operation of the SW or the criteria to be sent to the network; - connected to the preprocessor PE and the sensor set and the driver set SE, AT; a decoding logic network DIN that selects a sensor that has to supply data to the preprocessor PE or a driver that has to receive data and/or commands from the preprocessor PE; P.E.
2. The apparatus of claim 1, further comprising: a decoding logic network DIN; 7 Control logic of switching circuit SW CSW1,
to CSW2: - a manual control device for connecting the line units L1...L16 to the test reference generator DP;
A first bistable circuit S1, S2 that sends out during the operation of AS; AS1; AS2, which is a manual control device.
AS; AS1 and AS2 are line unit L1...
a first bistable circuit S1, in which said signal remains present until the connection between L16 and the network is restored;
S2; - Clock input is the first bistable circuit S1, S2
The data input is connected in parallel to the data input of the second bistable circuit of the control logic of all other line units; the output is connected to the data input via the first inverter IA1, IA2. a second bistable circuit D1, directly connected to the first input of the OR gates OR1, OR2;
D2, the output of the OR gate is connected to the switching circuit SW, and the second input 17-
1 is connected to the output of the driver that activates the switching circuit SW when a command arrives in the preprocessor PE; the reset input is connected in parallel to the reset input of the second bistable circuit of the control logic of all other line units. 7. The device according to claim 6, further comprising a second bistable circuit D1, D2 connected to the driver output via a second inverter IB1, IB2. 8 Telephone line 6 for connecting to line network RT
If (1)...6(32) is an analog telephone line, each line unit L1...L32 is further provided with: - a digital-to-analog conversion also connected to the output of the multiplexer MX and controlled by said clock signal; - connected to the output of the digital-to-analog converter CNA to limit the band of the converter's output signal to the telephone band and bring the converted signal to the level required for transmission over the analog telephone line 6; analog wave amplification system FA; - a third control connected to the output of the analog wave amplification system FA to detect the presence of a signal whose level is below a predetermined threshold when the converter CNA receives a silence coding signal; Circuit CA2
a third control circuit CA2, which sends out a logic signal indicative of a detected occurrence; - sends balanced audio signals to the analog telephone lines 6, decouples the DC signals of these lines from the audio, and decouples the DC signals of these lines from the audio; Another analog telephone line 15 is a transformer TR connected to the output of the wave amplification system FA for decoupling from the analog circuits FA, CA1, CA2, the output line of the transformer being connected to said switching circuit. 7. A device according to claim 6, comprising: a transformer TR for forming an audio loop; 9. Claims 7 or 8, wherein the second control circuit CA1 is connected to the output of the analog wave amplification system FA to detect the presence of a signal in the telephone band after conversion to analog form. The device described in any of the above. 10. The device according to claim 3, wherein each line unit L1...L16 is connected to two telephone lines and includes means for allowing access to the device via only one line at a time. . 11 If the two telephone lines are analog telephone lines, the audio loops of the two lines have the same sensors S, R ^* , TAX, BS and drivers A, A ^* , B,
^B
11. The device according to claim 10, wherein the device is connected to a driver D ^* that connects the voltage source 1 to a common voltage source. 12 The first input of the multiplexer MXa...MXn is a digital audio line 101a...101n dedicated to the line unit containing the multiplexer.
and the multiplexer connects all lines
Another digital audio line 3p...3 connected to the unit and dedicated to frequently requested announcements.
6. The apparatus according to claim 6, further comprising one or more other audio inputs respectively connected to the lines 3p...3r, wherein the announcement does not need to be heard from the beginning and is continuously sent to the lines 3p...3r. Device. 13 Digital audio lines 101a...10 connected to the first inputs of multiplexers MXa...MXn
1n is the output line of the switching circuit network RC,
The input of the above switching circuit network is the output line 101a.
...Digital line 3a, which is smaller in number than 101n...
Announcements and announcements from each audio unit U are made via 3m.
The audio lines 3p...3r connected to the first part Ua storing all announcements of the data base and connected to the other multiplexer inputs are the output lines of the second part Ub of said announcement data base. , the second part stores only announcements that are sent out continuously.
The device according to item 2.