Conference System

The present invention relates to a conference system which utilizes a time division digital switch.

Conventionally this kind of system employs a method which subjects participant's voices to an N-1 addition while holding their sound levels unchanged or after equally attenuating them for preventing singing. On this account, for example, in the case of a three-party communication involving a CO line A of a large line loss and intercom lines B and C of a small line loss, the intercom line B is supplied with the sum of voice (A) of the CO line A and voice (C) of the intercom line C, but the volume of the voice (A) is small and the volume of the voice (C) is large, so the voice of the CO line A is difficult to hear on the intercom line B.

Furthermore, when the participant's voices are equally attenuated for the prevention of singing, the voices that reach the participant of a large line loss are appreciably low and difficult to hear, and when a two-party communication is switched to a three-party one, sound volume variations are caused, creating a sense of incongruity.

It is therefore an object of the present invention to provide a conference system which is capable of freely setting the sound volume level by making the sound volume levels of the input and output of the conference circuit variable, for each participant, by pluralities of input and output ROM converters so that the levels of voices of participants of large and small line losses are made as close to each other as possible and that the participant of a large line loss is supplied with a voice of low attenuation.

The present invention will be described in detail below with reference to the accompanying drawings, in which:

FIG. 1 is a block diagram illustrating the general arrangement of a conference system to which the present invention is applied.

FIG. 2 is a block diagram illustrating the general arrangement of the conference system of the present invention;

FIG. 3 is a block diagram showing an example of a conference circuit for use in the present invention;

FIG. 4 is a timing chart showing the operation of the circuit exemplified in FIG. 3;

FIG. 5 is a timing chart showing the allocation of operation time in the circuit exemplified in FIG. 3;

FIG. 6 is a block diagram illustrating an example of a highway switching circuit for use in the present invention;

FIG. 7 is a timing chart showing the operation of the circuit exemplified in FIG. 6;

FIG. 8 (a) and (b) are diagrams showing memory maps of memories used in FIG. 3;

FIGS. 9 and 11 are charts explanatory of the circuit operation in FIG. 3;

FIG. 10 is a diagram showing imbodiments of conference trunk instructions indicating external interfaces in FIG. 3, and

FIG. 12 shows memory maps of memories in FIG. 6.

FIG. 1 shows the concept of the conference system to which the present invention is applied. A participant's signal provided on an input highway 1 is applied into a highway switching circuit 2 and a conference communication circuit 4 at the same time. In this instance, output switches 21, 22 and 23 of the highway switching circuit 2 are each turned OFF and output switches 41, 42 and 43 of the conference circuit 4 are each turned ON at the timing of the participants channel of the output highway. Furthermore, a signal conference-operated in the conference communication circuit 4 is output at the timing of the conferee's channel of the output highway. FIG. 1 shows, in the interest of clarity, a case where the number of conferees is three and their voice signals are identified by A, B and C, respectively, and are each assigned to a particular one of three channels of the highway.

As is evident from FIG. 1, since the participant's signal is not applied to the highway switching circuit 2 but is input directly into the conference circuit 4, the voice delay is smaller than in the past.

FIG. 2 represents the concept of the present invention, in which the participant's signal provided on the input highway 1 is stored in an input memory 11. A signal read out of the input memory 11 is applied to any one of table of first ROM converters 12, each determined by a participant, wherein it is converted from a PCM code to a linear code and at the same time its sound volume is attenuated (or amplified). The ROM converter output is provided to an operation circuit 13, wherein it is subjected to conference processing. Its output is applied to a second ROM converter determined by each participant, wherein it is converted from the linear code to the PCM code and at the same time its sound volume is attenuated (or amplified), thereafter being stored in an output memory 15. Accordingly, sound volume level of a conference signal which is delivered from the output memory 15 to the output highway 2 can be freely set in the first and second ROM converters 12 and 14 which are determined by each participant.

FIG. 3 illustrates an embodiment of the conference circuit 4 for use in the present invention, and designations of symbols used therein are given in Table 1.

The conference circuit 4 depicted in FIG. 3 is used, along with a highway switching circuit for controlling eight highways (32 channels/highway), and includes eight conference trunks (four participants/trunk).

The highway is divided into input highways and output highways, and though not shown, codecs corresponding to telephone sets or CO lines are each assigned the port number (the highway number+the channel number) by a fixed time slot system. Each codec performs transmission and reception of PCM data between it and the input/output highway at the timing of a channel determined by the port number. Further, PCM companding follows the .mu.-law.

A description will be given, with reference to FIG. 3, of the operation of the conference circuit. A counter (CNT) operates on external clock timings .phi. and frame pulses (FP), and timings T.sub.0 to T.sub.7 and timings S.sub.0 to S.sub.3 are created by a decoder (DEC.sub.2). The above timing relation is shown in FIG. 4. Data on an input highway (RHWY) from a codec of each port is converted by one of serial-parallel converters (HSP.sub.0 to HSP.sub.7) which operate on clock timings of .phi., and is transferred to one of highway registers (HWR.sub.0 to HWR.sub.7) at the timing T.sub.0.

To input memories (INM.sub.0 to INM.sub.7) are supplied addresses from the counter (CNT) via a multiplxer (MPX.sub.1) at the timing T.sub.7 and the data in the highway registers (HWR.sub.0 to HWR.sub.7) is stored into the input memories (INM.sub.0 to INM.sub.7) in such a format as shown in FIG. 4. Since the counter (CNT) operates as shown in the timing chart of FIG. 4, data of 32 channels is sequentially loaded into the input memories (INM.sub.0 to INM.sub.7) within one frame period 125 .mu.s. Of the data stored in the input memories (INM.sub.0 to INM.sub.7), data of conference-participating ports is subjected to conference processing described later and then stored in output memories (OTM.sub.0 to OTM.sub.7) corresponding to the conference-participating ports. The output memories (OTM.sub.0 to OTM.sub.7) are each supplied with, as an address, a value which is obtained by adding "2" in an adder (FA.sub.0) to a value applied thereto from the counter (CNT) via the multiplexer (MPX.sub.1) at the timing T.sub.7, and the corresponding data stored in one of the output output memories (OTM.sub.0 to OTM.sub.7) is transferred to one of parallel-serial converters (HPS.sub.0 to HPS.sub.7). On the other hand, when the output memories (OTM.sub.0 to TOM.sub.7) are read out, the same address as that applied to the output memories (OTM.sub.0 to OTM.sub.7) is also provided via a multiplexer (MPX.sub.3) to a control memory (CTM), at the same timing as the readout of the output memories (OTM.sub.0 to OTM.sub.7), by which the corresponding data in the control memory (CTM) is transferred to highway gate flip-flops (HWG.sub.0 to HWG.sub.7) on a bitwise basis. The control memory (CTM) has stored therein bits the number of which is equal to (the number of highways).times.(the number of channels), as shown in FIG. 8(2). The individual bits are independently set by an external processor in a manner described later. Accordingly, the highway gate flip-flops (HWG.sub.0 to HWG.sub.7) are set or reset according to the contents of the control memory (CTM), and only when they are in the set state, the outputs from the parallel-serial converters (HPS.sub.0 to HPS.sub.7) are transferred via bus drivers (BD) to output highways (THWY). When they are in the reset state, the outputs of the bus drivers (BD) are in a high-impedance state, in which they can be wired-ORed with the output of a highway switching circuit described later. The aforementioned conference processing of one trunk is executed, as shown in FIG. 5, in 15.625 .mu.s allocating to four channels which is obtained by equally dividing into one-eighth a transmission time 125 .mu.s of 32 channels. The process of operation concerning one trunk is shown in detail in the timing chart of FIG. 9 and its outline is as follows:

In a trunk memory (TKM) in FIG. 3, as depicted in FIG. 8(3), four words (15 bits/word) are assigned to one trunk, and each word has accommodated therein port information (the port number+ID) of four participants by a method described later. In this case, ID is individual information including a port loss of -0, a port loss of -1 and a connection flag, as shown in FIG. 8(3). To the trunk memory (TKM) is applied the output of the counter (CNT), as an address, via a multipexer (MPX.sub.0) at the timings T.sub.0 to T.sub.2 and T.sub.4 to T.sub.6. The address changes for each timing and four words of the same trunk are successively read out word by word at the timings S.sub.0 and S.sub.1, and at the timings S.sub.2 and S.sub.3, four words of the same trunk are read out again in succession. Thereafter the same operation is repeated for the next trunk. As a result, the above operation takes place for eight trunks in 125 .mu.s.

Now, when a first word (A) of a certain trunk of the trunk memory (TKM) is read out at the timing T.sub.0 to T.sub.2 of S.sub.0, speech data (a) is read out from that one of the input memories (INM.sub.0 to INM.sub.7) which corresponds to the port number (output at D.sub.0-4 and D.sub.5-7 of the trunk memory (TKM)), and is input into a terminal A.sub.0-6 of a multiplexer (MPX.sub.4). On the other hand, since the port loss of -1 (PL1) (derived at D.sub.8-10 of the TKM) included in the individual information (ID) of the word (a) is applied to a terminal A.sub.7-9 of the multiplexer (MPX.sub.4), the abovementioned speech data (a) is provided to that table of a .mu.-law to linear code translation table (MLT) in a read-only memory (ROM) which is specified by the port loss of -1, wherein it is converted from a PCM code (a) to a linear code (a') attenuated (or amplified) as predetermined, thereafter being delivered out from the read-only memory (ROM). The output of the read-only memory (ROM) is set in an area (ASP) of a speech register file (SPR) at the timing T.sub.2. The output of the area (ASP) is applied to an adder (FA.sub.1), wherein it is added with the output of an adding register (ADD), and the added output is stored in the adding register (ADD) at the timing T.sub.3.

In a similar manner, processing concerning second, third and fourth words in the aforementioned trunk is performed at the timings T.sub.4 to T.sub.7 of S.sub.0, T.sub.0 to T.sub.3 of S.sub.1 and T.sub.4 to T.sub.7 of S.sub.1. Now, let B, C and D stand for the respective words of (b, c and d) for speech data corresponding to the port numbers in the words of (B, C and D and b', c' and d') for values into which the speech data of (b, c and d) are converted by the read-only memory (ROM). At the point of completion of the timing T.sub.3 of S.sub.1, the values of (b', c' and d') are set in areas BSP, CSP and DSP of the speech register file (SPR), respectively, and the adding register (ADD) is cleared to zero at the timing T.sub.1 of S.sub.0, so a value of (a'+b'+c'+d') is stored in the adding register (ADD).

After this, the content (a') of the area (ASP) of the speech register file (SPR) is subtracted from the adding register (ADD) via the adder (FA.sub.1) at the timing T.sub.0 of S.sub.2, and as a result, the output (SPX) of the adder (FA.sub.1) goes to a value of (b'+c'+d'), which is input into a terminal B.sub.0-12 of the multiplexer (MPX.sub.4).

On the other hand, the word A is read out again from the trunk memory (TKM) at the timing T.sub.0 to T.sub.2 of S.sub.2, and since the port loss of -0 (PL0) (derived at D.sub.11-13 of the TKM) in the individual information (ID) of the word A is applied to B.sub.13-15 of the multiplexer (MPX.sub.4), the aforesaid speech data of (b'+c'+d') is provided to that table of linear to .mu.-law code translation table (LMT) in the read-only memory (ROM) which is specified by the port loss -0 (PL0), wherein it is converted from a linear code of (b'+c'+d') to a PCM code (b'+c'+d')' attenuated (or amplified) as predetermined, thereafter being output from the read-only memory (ROM). The output of the read-only memory (ROM) is loaded, at the timing T.sub.2 of S.sub.2, via a negative bus driver (NBA) into that one of the output memories (OTM.sub.0 to OTM.sub.7) which is specified by the port number in the word A.

Likewise, processing corresponding to the port numbers in the words B, C and D is performed at the timings T.sub.4 to T.sub.6 of S.sub.2, T.sub.0 to T.sub.2 of S.sub.3 and T.sub.4 to T.sub.6 of S.sub.3. At the end of each timing, codes (c'+d'+a')', (d'+a'+b')' and (a'+b'+c')' are stored in those of the output memories (OTM.sub.0 to OTM.sub.7) which correspond to the respective port numbers, thus completing the process concerning one trunk.

Since a processing time of 15.625 .mu.s is needed for one trunk, processing for eight trunks takes a processing time of 125 .mu.s.

As will be evident from the above, if the kinds of the .mu. law-to linear code translation table (MLT) for determining the input level and the linear to .mu. law code translation table (LMT) for determining the output level for each participant's port are set in the trunk memory in FIG. 3 at each time when the conference trunk is used, then the voice level of conference communication can suitably be altered according to the kind of the participants, such as a long distance line, a short distance CO line, and an intercom etc.

For example, in a memory map of the read-only memory (ROM) shown in FIG. 8(4), let it be assumed that x.sub.0, x.sub.1, x.sub.2, x.sub.3, x.sub.4, x.sub.5, x.sub.6 and x.sub.7 provide attenuations of 0, 1, 2, 3, 4, 5, 6 and 7 dB, respectively, and y.sub.0, y.sub.1, y.sub.2, y.sub.3, y.sub.4, y.sub.5, y.sub.6 and y.sub.7 attenuations of 0, -1, -2, -3, -4, -5, -6 and -7 dB (i.e. amplifications of 0, +1, +2, +3, +4, +5, +6 and +7 dB), respectively. In the case of three-party communication among a CO line A having a line loss of -4 dB and intercom lines B and C having a line loss of 0 dB, if different kinds of tables x.sub.0, y.sub.4 and x.sub.4, y.sub.0 are provided for the CO line A and the intercom lines B and C, respectively, voices of the intercom lines B and C are provided on the CO line A at 0 dB without attenuation (received at -4 dB at the far end of the CO line A), voices of the CO line A and the intercom line C are both provided on the intercom line B at -4 dB, and voices of the CO line A and the intercom line B are both provided on the intercom line C at -4 dB. Accordingly, volume of the same voice level as in a two-party communication between the intercom line B or C and the CO line party A can be obtained in the three-party communication, and since no level difference exists between the CO line and the intercom lines, voices are easy to be hand. On the other hand, volume on the CO line A during the three-party communication is also at the same level as in the case where the CO line A is in the two-party communication with the intercom line. Therefore, switching from the two-party to the three-party communication does not create a sense of incongruity.

While in the above the code translation tables are set so that no difference in volume is present between the intercom line B or C and the CO line A or the intercom line C or B, it is evident that the code translation tables can be set in a manner to provide a difference of several dB between them and that voices can be amplified for the CO line with a large line loss, and the combination of code translation tables can be changed according to a particular purpose. Furthermore, it is also possible to employ different levels for two-party, three-party and four-party communications, respectively.

Memory maps of the trunk memory (TKM) and the control memory (CTM) are as shown in FIGS. 8(3) and (2), and required data is set in them by an external processor using output instructions depicted in FIG. 10. This is shown in detail in the timing chart of FIG. 11 (wherein the arrows indicate destinations), and its outline is as follows: An output command OUT30 sets the number of a conference circuit (an IC number), the trunk number in the conference circuit and the position number in the trunk in a trunk number register (TNO), an output command OUT31 sets the participant's port number in a port number register (PNO) and then an output command OUT32 sets a connection flag to a "1" in a connection information register (CON). If a comparator (COMP) detects coincidence between the IC number set in the trunk number register (TNO) and an externally provided IC number, the output of the trunk number register (TNO) is applied to the multiplexer (MPX.sub.0) at the timing T.sub.3 or T.sub.6 to produce an address of the trunk memory (TKM), and the contents of the port number register (PNO) and the connection information register (CON) are loaded via a bus driver (BD) into the trunk memory (TKM) at the position of a specified trunk number in such a format as shown in the memory map in FIG. 8(3).

On the other hand, the channel number in the output of the port number register (PNO) is applied to the multiplexer (MPX.sub.3) to provide an address of the control memory (CTM), and data corresponding to a specified channel of the control memory (CTM) is set in a control memory register (CTR) at the timing T.sub.1. The output of the control memory register (CTR) is applied to a bit change circuit (CHG), wherein its only one bit specified by the highway number in the output of the port number register (PNO) is changed to a "1". The output of the bit change circuit (CHG) is stored again via a bus driver (BD) in the control memory (CTM) at the aforesaid channel position at the timing T.sub.2 or T.sub.6.

The above has described the setting of one port in the trunk memory (TKM) and the control memory (CTM), and the operation for its resetting is identical with the above except in that the connection flag of the output command OUT32 is made state of a "0".

For setting or resetting a plurality of ports, it is necessary only to repeatedly execute the output command OUT30, OUT31 and OUT32.

According to the above embodiment of the conference circuit, a maximum eight of such circuits can be connected in parallel, permitting the provision of a conference trunk having up to 64 trunks.

It is apparent that up to 256 trunks can be obtained by a two-bit increase in the bit lengths of the trunk number register (TNO) and the output instructions OUT30.

FIG. 6 illustrates an example of the highway switching circuit 2 for use with the conference circuit 4 in the present invention. Table 2 shows the designations corresponding to symbols of its respective parts. The highway switching circuit depicted in FIG. 6 controls eight highways (32 channels/highway), each divided into input and output highways 1 and 3. Codecs corresponding to CO lines or telephone sets, though not shown, are each assigned the port number (the highway number+the channel number) through the fixed time slot system, and each codec conducts transmission and reception of PCM data between it and the input highways and the output highways at the timing determined by the port number.

A description will be given, with reference to FIG. 6, of the operation of the highway switching circuit 2. A counter (CNT) operates on an external clock timings of .phi..sub.0 and a decoder (DEC) creates timing slot T.sub.0 to T.sub.7, S.sub.0, S.sub.1, u.sub.0 to u.sub.7 and FP, as shown in FIG. 7. The timing slots T.sub.0 to T.sub.7 are to identify the highway numbers, the timing S.sub.0 is assigned to the data input from the input highway (RHWY) to a speech data memory (SPM) and the signal S.sub.1 is assigned to the data output from the speech data memory (SPM) to the output highway (THWY). Data on the input highway (RHWY), which is sent from the codec of each port number, is converted by one of serial-parallel converters (SP.sub.0 to SP.sub.7) which operate on a clock timings of .phi..sub.1, and is transferred to one of input registers (INR.sub.0 to INR.sub.7) at the end of the timing u.sub.7 (i.e. at the beginning of the timing u.sub.0). To the speech data memory (SPM) is applied an address from the counter (CNT) via the multiplexer (MPX.sub.1) at the timing S.sub.0. Since this address is updated upon each occurence of the clock timings of .phi..sub.0, the data in the input registers (INR.sub.0 to INR.sub.7) is sequentially loaded into the speech data memory (SPM) via bus drivers (BD.sub.0 to BD.sub.7) selected by the timing slots T.sub.0 to T.sub.7, as shown in FIG. 12(1). In FIG. 12(1), reference characters CHm and HWn indicate speech data which is input from the port corresponding to an mth channel of an nth highway, and each data is composed of a + or - sign (S), a chord and a step. On the other hand, a connection memory (CONM) is supplied with an address from the counter (CNT) via the multiplexer (MPX.sub.0) at the timing S.sub.1. Of the contents of the connection memory (CONM) shown in FIG. 12(2) (wherein CHm and HWn indicate, respectively, the port number and an output control flag of the destination from which the speech data is delivered to the port corresponding to the mth channel of the nth highway), read out by the above address, the port number of the destination is provided as an address via the multiplexer (MPX.sub.1) to the speech data memory (SPM). The speech data read out by the address from the speech data memory (SPM) is successively loaded into output registers (OUTR.sub.0 to OUTR.sub.7) which are each selected by one of the timing signals T.sub.0 to T.sub.7 upon each occurrence of the clock timings of .phi..sub.0. The contents of the output resisters (OUTR.sub.0 to OUTR.sub.7) are stored in parallel-serial converters (PS.sub.0 to PS.sub.7) at the end of the timing signal u.sub.7 (i.e. at the beginning of the timing signal u.sub.0) Of the contents of the connection memory (CONM), the output control flags are directly loaded in the output registers (OUTR.sub.0 to OUTR.sub.7) at the timings T.sub.0 to T.sub.7 and thence loaded in flip-flops (FF.sub.0 to FF.sub.7) at the timing u.sub.0. The parallel-serial converters (PS.sub.0 to PS.sub.7) operate on the clock timings of .phi..sub.1 and their outputs are provided on the output highway (THWY) via gates (G.sub.0 to G.sub.7) when the flip-flops FF.sub.0 to FF.sub.7) are in their set state, but when the flip-flops (FF.sub.0 to FF.sub.7) are in their reset state, the outputs of the gates (G.sub.0 to G.sub.7) are in the high-impedance state, permitting the output of the conference circuit 4 to be wired-ORed.

Incidentally, frame pulses (FP) from a decoder (DEC) are provided to the conference circuit 4 for putting the channels of the conference circuit 4 and the highway switching circuit 2 in phase with each other. A writing operation of information in the connection memory (CONM) is effected by an external processor via an external interface circuit through the use of an output instruction at the timing S.sub.0. This is accomplished by the same method as described previously in connection with the conference circuit 4, and hence will not be described.

The above is an embodiment of the present invention and should not be construed as limiting the invention specifically thereto. For example, the number of participants in one trunk may be three or five. The number of trunks used may also be determined. Furthermore, the number of highways used need not always be limited specifically to eight. While in the embodiment the input highways and the output highways are physically isolated from each other, it is evident that they can be isolated in terms of time but, physically, can be used so as to have both input and output highways functions. Moreover, it is also easy to employ, as the PCM code, an A-law code in place of the .mu.-law code.

In short, various modifications may be effected without departing from the scope of the gist of the present invention. Hence the invention can be applied widely to various communication apparatus.

As described above, the present invention permits free and dynamic setting of the volume level of a conference with very simple control, and hence is advantageous in that the volume difference between a CO line and an intercom line can be reduced or volume variations can be suppressed when switching a two-party communication to a three-party one, i.e. the volume level can be varied according to a particular purpose.