Switching system having capability for telecommunication conference.

BACKGROUND OF THE INVENTION

(1) Field of the Invention

The present invention relates to a switching system, more particularly, to a technique for a telecommunication conference in a digital switching system incorporated in a time division multiplex network.

(2) Description of the Related Art

Generally, many subscribers are accommodated in a digital switching system. Conversations are usually performed between two subscribers in' accordance with switching control. Recently, however, switching systems are being designed to enable telecommunication conferences wherein conversations are performed between three, four or more subscribers. A telecommunication conference is achieved relatively easily by a digital switching system operated in a time division multiplex network through a mixing process with a circuit switching. With the above-mentioned mixing process, voice signals transmitted from.a plurality of subscriber terminals (telephone set) are mixed and then distributed to the subscribers concerned.

In a prior art switching system, the mixing process is achieved based on a trunk (TRK).

Therefore, two problems arise. First, if the trunk is designed as a trunk for three persons, the trunk is necessarily limited to use by no more than three subscribers for a conference. Similarly, if it is designed as a trunk for four persons, the trunk is necessarily limited to use by no more than four subscribers. When a trunk is required to cope with another three or more subscribers, the related hardware, such as the incoming-buffer registers and outgoing-buffer registers, necessarily becomes large in scale.

Second, a suitable number of the mixing trunks has to be always available so as to deal with traffic conditions in the switching system, which conditions are determined, in advance of when the switching system is built.

SUMMARY OF THE INVENTION

An object of the present invention is to provide a switching system capable of overcoming the two problems simultaneously. That is, the switching system of the present invention enables telecommunication conferences among any number of subscribers with simple hardware.

To attain the above object, the switching system of the present invention features primarily a mixing circuit and a feedback circuit instead of the prior art mixing trunk. The mixing circuit operates to mix two voice data signals read from a speech path memory (SPM). The thus mixed voice data signals are rewritten, with the aid of the feedback circuit, into the speech path memory.

BRIEF DESCRIPTION OF THE DRAWINGS

The above object and features of the present invention will be more apparent from the following description of the preferred embodiments with reference to the accompanying drawings, wherein: <ul id="ul0001" list-style="none"><li>Fig. 1 is a schematic representation of a mixing process for a telecommunication conference;</li><li>Fig. 2 is a general view of a switching system in which a conventional mixing operation is achieved;</li><li>Fig. 3 illustrates in detail an example of the mixing trunk shown in Fig. 2;</li><li>Fig. 4 is a switching system according to an embodiment of the present invention;</li><li>Figs. 5(a) through 5(d) depict time charts used for explaining a general operation achieved between the speech path memory, and the speech path control memory, both shown in Fig. 2;</li><li>Fig. 6 illustrates part of a modification of the switching system of Fig. 4;</li><li>Fig. 7 is a schematic layout view of both the speech path memory (SPM) and the speech path control memory (SCM) for advantageously achieving the mixing operation in the switching system according to the present invention;</li><li>Fig. 8 is one example of a known digital pad;</li><li>Fig. 9 is a detailed example of the mixing circuit sown in Fig. 4;</li><li>Fig. 10 illustrates part of the modification of the switching system of Fig. 4, which employs a direct feedback means of voice data signals therein;</li><li>Fig. 11 is a detailed example of the mixing circuit shown in Fig. 10;</li><li>Fig. 12(a) through 12(e) depict timing charts used for explaining the mixing operation according to the present invention;</li><li>Figs. 13(a), 13(b), 13(c), 13(a') 13(b'), and 13(c') depict timing charts used for explaining in detail the present invention by taking a case where four subscribers hold a conference as an example; and</li><li>Fig. 14 illustrates a general construction of the timing control unit.</li></ul>

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Before describing the embodiments of the present invention, the prior art and the problems therein will be first described with reference to the related figures.

Figure 1 depicts a schematic representation of a mixing process for a telecommunication conference. In this figure, reference numeral 11 indicates a plurality of subscriber terminals, such as telephone sets. Any two or more subscriber can hold conferences or conversations with each other by way of a switching system (SWS) 12 to which the present invention refers. During a telecommunication conference with the subscribers, A, B, C, and D, for example, the switching system 12 receives the voice data signals V_A , V_B, V_C , and V_D given from the respective subscriber terminal (11) and then applies mixing operation to the thus received voice data signals, so that the telephone set 11A is supplied with the mixed voice data signals V_B + V_c + V_D , excluding the own side signal V_A. Similarly, the telephone sets 11B, 11C, and 11D are supplied with respective mixed voice data signals, each excluding own side signal, as represented by characters of the voice data signals (V) along the respective subscriber lines SL. Thus, each subscriber can hear voices from other subscribers simultaneously. As a result, the related subscribers can hold a telecommunication conference.

The telecommunication conference is performed, in the prior art, with the mixing operation explained below. Figure 2 is a general view of a switching system in which a conventional mixing operation is achieved. In Fig. 2, members identical to those previously described are represented by the same reference numerals or characters (same for later figures). In Fig. 2, the prior art means for carrying out the mixing operation is characterized by a mixing trunk (MXT) 23. The telephone sets llA, llB, 11C, 11D, and so on are connected with a multiplexer circuit (MUX) 22 by way of individual subscriber circuits 21A, 21B, 21C, 21D, and so on. There are, in actuality, many telephone set (11) and subscriber circuit (21) pairs, however, only four pairs, i.e., 11A-21A through 11D-21D, are illustrated for simplicity. Primary constituents of each subscriber circuit (21) are, as illustrated, a two-wire/four-wire system conversion circuit, i.e., a hybrid circuit HYB, a coder COD mounted at a transmitter side, and a decoder DEC mounted at a receiver side. For example, pulse code modulation (PCM) coded voice data signals are transmitted from the coder COD and applied to the multiplexer circuit 22 serially one by one for every telephone set (11). The thus applied serial PCM coded voice signals are converted into parallel voice data signals by the multiplexer circuit 22, which is known as a time division multiplex. Each frame of the thus multiplexed, in time division, voice data signals of, for example, eight bits is sequentially written at respective allotted time slots into each memory element of a speech path memory (SPM) 24 via an input bus BS_in and thus the 8-bit voice data signals are stored in the memory 24.

On the other hand, a read operation from the memory 24 precedes the above-mentioned write (store) operation therein. The read operation is achieved in accordance with address information given by a speech path control memory (SCM) 25. Therefore, the voice data signals stored in the memory 24 are read therefrom and output to a first output bus BSoutl. Incidentally, a read-after-write method may also be used for the read/write control of the speech path memory 24, instead of the above-mentioned write-after-read method.

Regarding the speech path control memory 25, which voice data signals of the memory elements should be specified and read out therefrom is determined in accordance with correlation to individual time slots to which the subscriber is allotted, should hold a telecommunication conference or conversation. In the speech path control memory 25, many memory elements are set up for each time slot, as is in the speech path memory 24, so as to store therein the address data for specifying the voice data signals to be read from the speech path memory 24. The gist of the address data is determined by central control equipment (CC) 26 in accordance with dial information sent from an originating subscriber and received, by way of a scanner, thereby.

The voice data signals applied to the first output bus BS, are transferred to a second output bus BS₂ via a digital pad (PAD) and given to a demultiplexer circuit (DMX) 28 by which the thus given voice data signals are demultiplexed for every time slot o, j, k, L, and so on, to produce transmitting voice data signals, such as represented by a, b, c, d, and e. These transmitting voice data signals a, b, c, and so on are received by the subscriber circuits (21) allotted with respective time slots and then reach each subscriber unit, i.e., the telephone sets (11), by way of individual decoders DEC.

The operations for performing the above-mentioned switchings are carried out in time slot units. Accordingly, timing management for time slots is required. For this, a timing control unit (TG) 29 is employed in the system. Thus, the switching system can work as a so-called time division multiplex network.

In the prior art switching system concerned, the mixing operation is achieved with the use of the mixing trunk (MXT) 23, as mentioned previously. Figure 3, illustrates in detail an example of the mixing trunk shown in Fig. 2. Supposing here that a telecommunication conference is to be held with three subscribers, first, the voice data signals concerned, corresponding to respective time slots, are given, as the transmitting voice data signals e, from the demultiplexer circuit 28 and next applied to corresponding incoming buffer memories (INBR₀, INBR_-, INBR₂) 31-0, 31-1, and 31-2 for momentarily storage therein. The stored voice data signals are supplied to one input of an adder (ADD) one by one with predetermined timings selectively by means of a selector (SEL).

Under control of the timing control unit (TG) 29, first, a register (REG) 35 is cleared to hold an initial value 0. Second, the gist of the incoming buffer register (INBR_O) 31-0 is input, by control of the selector (SEL) 33, to the adder 34. As a result, the gist of the incoming buffer register 31-0 is set in the register 35. Next, the incoming buffer register 31-1 is selected by the selector 33 and the gist thereof is supplied to one input of the adder 34. At the same time, the gist, set now in the register 35, is supplied to the other input of adder 34 to add the related two voice data signals at its two inputs. The resultant add signal from the adder^*34 is set in the register 35. Further, the incoming buffer register 31-2 is selected by the selector 33 and the gist thereof is supplied to one input of the adder 34. At the same time, the gist, set now in the register 35, is supplied to the other input of the adder 34 to add the related two voice data signals at its two inputs. Thus, the register 35 contains therein the gists of the incoming buffer registers 31-0, 31-1, and 31-2. These gists comprise information being communicated by the three subscribers which hold the conference.

After accumulation of the three voice data signals in the register 35, under control of the timing control unit (TG) 29, the adder 34 is converted to act as a subtractor. It receives each gist from the incoming buffer registers 31-0, 31-1, and 31-2 sequentially by the control of a selector 33. The gist of the register 35 is subtracted by the gist of the incoming buffer register 31-0 and the thus subtracted gist is transferred to the outgoing buffer register (OUTBR_O) 36-0. During the above-mentioned subtraction, the gist of the register 35 is not renewed. Therefore, the register 35 holds therein the mixed telecommunication information as it is and supplies each bit of telecommunication information excluding each own side information sequentially to the outgoing buffer registers (OUTBR₁ , OUTBR₂) 36-1 and 36-2, The gists thus supplied to the outgoing buffer registers 36-0, 36-1, and 36-2 are fed back to the multiplexer circuit 22 (Fig. 2) and finally distributed to the three subscribers concerned by way of the speech path memory 24, the digital pad 27, and the demultiplexer circuit 28. Consequently, the outgoing buffer registers 36-0, 36-1, and 36-2 store therein the voice data signals V_B+ V_C , VA + V_C, and V_A+ V_B, respectively, assuming that the incoming buffer registers 31-0, 31-1, and 31-2 have the three subscribers voice data signals V_A , V , and V of Fig. 1, respectively.

Figure 4 is a switching system according to an embodiment of the present invention. The differences here from Fig. 2 are that, first, no mixing trunk 23 is used and, second, a mixing circuit (MIX) 41 is employed and mounted between the digital pad 27 (or the speech path memory 24 if such pad 27 is not used) and the demultiplexer circuit 28. The mixing operation itself for the voice data signals is quite different from the prior art mixing operation, although the related difference is not well expressed in the figure. It should be understood, however, that the basic switching operation is substantially the same as that explained with reference to Fig. 2. Figures 5 (a) through 5 (d) depict time charts used for explaining a general operation achieved between the speech path memory and the speech path control memory both shown in Fig. 2. The time charts of Fig. 5 well exemplify the above-mentioned basic switching operation carried out in not only the prior art switching system of Fig. 2, but also the switching system of the present invention shown in Fig. 4. It should be noted that the speech path memory (SPM) 24 may be directly controlled by the central control equipment (CC) 26 without introducing a speech path control means therebetween, as shown in Fig. 6. Figure 6 illustrates a part of a modification of the switching system of Fig. 4. This modification can also be applied to the switching system of Fig. 2.

The time charts of Figs. 5(a) through 5(d) represent arbitrary one frame FR extracted selectively from an (n-1) th time slot TS_n-1 to an (n+2)th time slot. To be specific, Fig. 5(a) depicts input voice data signals, represented as VI₁ and VI₂ , to be written into the speech path memory (SPM) 24 and provided on the input bus BS_in therefor, Fig. 5(b) indicates operation modes of the speech path memory (SPM) 24, represented as a read operation mode R and a write operation mode W, Fig. 5(c) depicts address data given from the speech path control memory (SCM) 25, if it is used, represented by A₁ and A₂, and Fig. 5(d) depicts output voice data signals to be read from the speech path memory (_SPM) 24 and supplied to the first output bus BS_outl, represented as VO₁and V0₂.

As seen from Fig. 5(b), the speech path memory (SPM) 24 works to achieve both the read operation and the write operation alternately for every time slot. The read operation reads the voice data signals which have already been written into the memory 24 at the corresponding time slot in the frame FR appearing immediately before. The write operation deals with the voice data signals represented by VI₁ and VI₂. The read operations are achieved in accordance with the address data, such as A₁ and A₂, given from the speech path control memory (SCM) 25, whereby the voice data signals of the desired remaining side subscribers can be read from the memory (SPM) 24. On the other hand, the input voice data signals, such as VI₁ and VI2 , are automatically sequentially written into individual memory elements, each corresponding to the allotted time slot, of the speech path memory 24. Further, each frame FR has a period of, for example, 125 µs in time and is composed of, for example, 256 time slots, i.e., TS_O through TS₂₅₅. In this case, these time slots TS_O through TS₂₅₅ are allotted, as a rule, to the subscribers numbered as "0" through "255", respectively. If lesser subscribers than that number (256) are accommodated in the switching system concerned, some time slots may be unused. It should be noted, however, that such unused time slots are important constituents of the present invention to perform the intended mixing of the voice data signals, which will be clarified hereinafter.

Figure 7 shows a schematic layout of both the speech path memory (SPM) and the speech path control memory (SCM) for advantageously achieving the mixing operation in the switching system according to the present invention. The layout is directed to a case where the mixing operation is performed among, for example, four subscribers (A, B, C, and D), as in the case of Fig. 1. It should be noted, however, that identical layout can also be applied to other cases, such as where three subscribers and five subscribers need the same mixing operation. This will be clarified by the following explanations. Further, it will consequently be recognized that the present invention features flexible application of the mixing operation to any member of subscribers. In Fig. 7, the left column corresponds to the speech path memory (SPM) 24 in which the digital voice data signals V are stored. The right column corresponds to the speech path control memory (SCM) 25 into which the address data A, used for accessing the speech path memory 24, are written. In other words, information for switching the time slots, i.e., line switching information, are written into the memory 25 in accordance with the related instructions issued from the central control equipment (CC) 26. Also stored therein is control information P, which is to be given to the digital pad (PAD) 27. The digital pad 27 functions, on one hand, to apply large attenuation to the voice data signals communicated between, for example, the subscribers A and B, i.e., the telephone sets llA and 11B, both accommodated in the same office, and, on the other hand, to apply small attenuation to the voice data signals communicated between one subscriber of one office and another of a remote office. This enables equalization of all the voice levels. The equalizations in voice levels are important for actual telecommunication conferences, however, are not so essential for the present invention.

Figure 8 is one example of the known digital pad. The digital digital pad 27 is chiefly fabricated with a read-only memory (ROM). The ROM works to produce output codes each containing"the respective aforesaid attenuation information. To be specific, the ROM is accessed by addresses each comprised of both the 8-bit voice data signal from the SPM and the corresponding 2-bit attenuation specifying signal.

Returning again to Fig. 7, particular attention should be paid to the following two facts. First, the memory elements ME's are considered sectioned into a memory area MA1 for the line switching and memory area MA2 for the mixing operation in each of the memories (SPM, SCM) 24 and 25. In this case, the memory areas MA1 for the line switchings per se exist also in the prior art memories (SPM, SCM) 24 and 25 and are inherent thereto. Therefore the memory areas MA2 for the mixing operations play an especially important role in the present invention.

Second, in the memory area MA2 for the mixing operation of the speech path control memory (SCM) 25, the memory elements set up mixing pairs MXP's. This results in each mixing operation being applied to the voice data signals contained in each pair of the two successive time slots. The characters, as shown by j, k, i, nO, nl, and so on in Fig. 7, represent respective addresses per se or address data and are allotted to respective time slots. When the address data is given from the speech path control memory (SCM) 25, the voice data signal is read out from the speech path memory (SPM) 24 at its address corresponding to the thus given address data. The thus read voice data signal is supplied to the mixing circuit (MIX) of Fig. 4 via the digital pad 27.

Figure 9 is a detailed example of the mixing circuit shown in Fig. 4. As seen from Fig. 9, the mixing circuit (MIX) 41 is comprised of a register (REG) 42, an adder (ADD) 43, and a selector (SEL) 44. The register 42 operates to momentarily store the voice data signal provided via the output bus BS_out2 from the digital pad 27 for every time slot. The thus momentarily stored voice data signal in the register 42 is digital added, in the adder 43, with a next supplied digital voice data signal appearing at the time slot next to the preceding time slot. The selector 44 operates to choose either one of the two inputs selectively and transfer the chosen one to the demultiplexer circuit (DMX) 28 via the'output bus BSout3. The two inputs are, first, the voice data signals after the adder operation at the adder 43 and, second, the usual voice data signals to which no such adder operations are effected. The operations mentioned above are controlled in timing, as a whole, by the timing control unit (TG) 29, which controls the selector 44 to choose the signals of the adder 43 side at every time slot for the mixing operations.

The central control equipment (CC) 26 performs a usual switching control operation if no telecommunication conference is held. Once a telecommunication conference is requested by subscribers, the central control equipment 26 operates to allot time slots therefor and to issue a command to the speech path control memory (SCM) 25 to use the thus allotted two slots for the mixing operations.

As may be understood from the preceding explanations, the present invention is characterized by the mixing circuit (MIX) 41 and the feedback means. The feedback means is operative to feed the mixed voice data signals back to the speech path memory (SPM) 24. It should be noted that the feedback means mentioned before is formed as an indirect feedback type. To be more specific, the feedback means is realized by the demultiplexer circuit (DMX) 28 and the multiplexer circuit (MX) 22, as understood with reference to Fig. 4, though the actual connecting lines for the feedback are not illustrated therein. On the other hand, it should be noted that identical feedback means can be formed of by a direct feedback type, as explained below.

Figure 10 illustrates part of a modification of the switching system of Fig. 4, which employs a direct feedback means for voice data signals. The direct feedback means is realized with feedback lines 101 and a switching gate 102 with a mixing circuit (MIX) 41' inserted therein. The mixed voice data signals are accumulated in the speech path memory (SPM) 24 without passing through the multiplexer and demultiplexer circuits 22 and 28. The indirect feedback type telecommunication conference switching system is suitable for a relatively small capacities. Since the multiplexer circuit 22 is not used in the system of Fig. 10, the switching gate 102 is required, which is comprised of, for example, a first driver 103 and a second driver 104 inserted in the feedback line 101 and the input bus BS. respectively, and controlled by the timing control unit (TG) 29 so as to energize or deenergize the same selectively. Usually the second driver 104 is energized, but the first driver 103 is energized every time the output from the mixing circuit 41' is provided. The construction of the mixing circuit 41' is slightly different from that of the mixing circuit 41 illustrated in Fig. 9.

Figure 11 shows a detailed example of the mixing circuit shown in Fig. 10. The difference have relative to the construction of Fig. 9 is that the route bypassing the adder 43 of Fig. 9 is not used.

Figures 12(a) through 12.(e) depict timing charts used for explaining the mixing operation according to the present invention. The timing charts are viewed from the mixing circuit 41 or 41' (Figs. 4, 9 and 11). Figure 12(a) relates to the voice data signals provided to the output bus BS_out2, i.e., the input of the mixing circuit 41 or 41', Fig. 12(b) to each set pulse S_P given from the timing control unit (TG) 29, by which set pulses SP the voice data signals are sequentially introduced into the register (REG) 42, Fig. 12(c) to the voice data signals thus introduced into the register 42 by each set pulse SP, Fig. 12(d) to a switching state determined by the selector (SEL) 44 or the switching gate 102 (Fig. 10), wherein the "0" state denotes a case where the voice data signals are transferred without passing through the adder (ADD) 43 (if the switching gate 102 is used, the second driver 104 is energized) and "1" state denotes a case where the voice data signals are transferred by way of the adder 43 (if the switching gate 102 is used, the first driver 103 is energized), and Fig. 12(e) to the outputs from the adder 43 and appearing on the output bus BS_out3 or the feedback line 101 (if the system of Fig. 10 is used). It should be understood that Figs. 12(a) through 12(e) refer to an arbitrarily selected one of the frames FR's. Each of the frames FR's is sectioned into the time slot train TSEX for the line switching and the time slot train TSMX for the mixing operation, thus sectioning also characterizing the present invention. In this case, the time slot train TSEX is allotted for usual voice data signal switching, i.e., the signals which need no mixing operation. A detailed explanation of the time slot train TSEX for usual line switching is omitted, since the time slot train TSEX is not essential to the present invention.

The time slot trains TSEX and TSMX are allotted to the aforesaid memory area MA1 for the line switching and the memory area MA2 for the mixing operation (both shown in Fig. 7). Therefore, the mixing pair MXP in the time slot train TSMX for the line switching corresponds to the aforesaid mixing pair MXP of the speech path control memory 25 shown in Fig. 7, wherein the mixing pairs MXP's in Fig. 12(a) are specifically illustrated by each two successive time slots, such as "TS_r+2, TS_r+3" and "TS_r+4' TS_r+5".

Here, characters I, II, III, IV, X, and XI are used to simply distinguish the voice data signals from each other, as shown in Fig. 12(a). The voice data signals I, III, and X are located at the head side in the individual mixing pairs MXP's, as depicted in Fig. 12(a), and are picked up by respective set pulses SP's of Fig. 12(b) to produce the voice data signals I, III, and X to be stored in the register (REG) 42, as shown in Fig. 12(c). Then, the thus stored signals I, III, and X are added, in the adder 43, to the signals II, IV, and XI appearing after these signals I, III, and X, respectively, to produce the resultant mixed voice data signals "I + II", "III + IV", and "X + XI" as shown in Fig. 12(e), respectively. The mixed voice data signals are, at last, accumulated in the speech path memory 24. At this time, nonused time slots 0 are necessarily left for every one time slot, as shown in Fig. 12(e).

The outputs of the adder 43, i.e., the mixed voice data signals, are send from the demultiplexer circuit (DMX) 28, at its respective output terminals, as the output e of Fig. 4 and then input to the multiplexer circuit (MX) 22, at its respective input terminals, as the identical input e. The thus input voice data signals are fed back to the memory area MA2 of the speech path memory (SPM) 24 shown in Fig. 7 and stored in a specified memory element ME. The address of the specified memory element ME is allotted to the time slot to which the related subscriber is allotted, as in usual line switching. Thus, the mixing of the voice data signals is achieved for each two subscribers, first, and the thus mixed voice data signals are accumulated in the memory area MA2 of the SPM 24. The mixing and accumulation operation is repeated as follows: "V_A⁺ V_B" accumulation is (sufficient accumulation for mixing for three subscribers) + "V_A+ V_B + V_C" (sufficient accumulation for mixing for four subscribers) "V_A + V_B + V_c + V_E" (sufficient accumulation for mixing for five subscribers, same for later explanation), and "V_C+ V_D" → "V_B + V_C+ V_D". As a result, the desired mixed voice data signals are prepared for the telecommunication conference. The plurality of sets of these mixed voice data signals are prepared in a similar manner during each frame FR and will be distributed during the next frame FR in the form of "V_A⁺V_B + V_C", "^VB ⁺ V_C+ V_D", and the like, as illustrated in Fig. ₁. A more detailed explanation will follow with reference to Figs. 13(a), 13(b), 13(c), 13(a'), 13(b'), and 13(c').

Figures 13(a), 13(b), 13 (c), 13(a'), 13(b'), and 13(c') depict timing charts used for explaining in detail the present invention, taking the case of four subscribers holding a conference as an example. The timing charts of these figures are based on the layout exemplified in Fig. 7 for easy comprehension. Therefore, the following explanations will be made with reference to Figs. 4, 7, 9, and 12(a) through 12(e). Figures 13(a) and 13(a') show two parts of a single line. The same applies to Figs. 13(b), 13(b') and Figs. 13 (c), 13 (c'). Figures 13(a) and 13(a') relate to the voice data signals transferred on the input bus BS_in , Figs. 13(b) and 13(b') to the voice data signals on the output bases BS_out2 and BS_out3' and Figs. 13 (c) and 13 (c') to the voice data signals on the output bus BS_out3. In this case, the voice data signals on the bus BS_out3 are assumed to be the same as those on the bus BSout2 for the sake of simplicity. That is, no adjustment, by the digital pad 27, in level of the voice data signals is executed. Further, for the sake of further simplicity, no time delay caused by the circuit elements is taken into consideration. Furthermore, the frame, shown as FR commonly in Figs. 13(a) through 13(c'), is sectioned, as previously mentioned, into the time slot train TSEX for the line switching and the time slot train TSMX for the mixing operation, which time slot trains are composed of, as a whole, the time slots TS₀, TS_j , TS_k --- TS_n0 , TS_n1 --- TS_n11. The suffixes such as j, k, n0, nl, n11,--- accompanying each character TS, are allotted to the addresses j, k, n0, nl, nll, --- for accessing the speech path control memory (SCM) 25. The suffixes j, k, n0, nl, nll --- are also allotted to the address data j, k, n₀, nl, nll --- which are stored in the speech path control memory (SCM) 25. In the operation, first, the voice data signals appear, such as shown by VA , V_B, V_C --- in front part of Fig. 13(a), and are then written into the speech path memory (SPM) 24 at the corresponding addresses j, k, ℓ ---, respectively. The write operation is represented by characters W·SPM in the related drawing. The thus written voice data signals are then supplied to the output buses BS_out1 and BS_out2, as depicted in the front part of Fig. 13(b'), this being done by the read operation in the speech path memory 24, which read operation is represented by characters SPM·R. The above operations will be clarified along with the example shown in Fig. 7. According to the example of Fig. 7, the address data j is stored in the speech path control memory 25 at its address n0, corresponding to the time slot TS_n0 (Fig. 13), of the memory area MA2 for the mixing operation. Therefore, the gist at the address j in the speech path memory 24 is read therefrom as, in this case, the voice data signal V_A and then supplied to the output buses BS_out1 and BS_out2, In a similar way, these buses BS_out1 and ^BS_out2 are supplied with the voice data signals V_B , V_C --- one by one, as depicted in the front part of Fig. 13(b'), and then applied to the mixing circuit (MIX) 41. In the mixing circuit 41, the aforesaid add operations are executed at the adder 43 to produce the mixed voice data signals in the form of "V_A + V_B", which will then be provided to the output bus ^BS_out3' The add operation is represented by characters ADD in the related figure.

The mixed voice data signals are fed back, by way of the demultiplexer circuit (DMX) 28, to the multiplexer circuit (MX) 22 and appear again on the input bus BS_in in the form of "V_A + V_B" at the time slot TS_n1 as depicted in the front part of Fig. 13(a'). The signals "V_A + V_B" concerned are again written into the speech path memory 24 at its addresses nl as illustrated by "V_A + V_B" of the memory 24 in Fig. 7. This completes the first stage accumulation of the voice data signals. It should be understood that, regarding the transition of the operation from the front part of Fig. 13(c') to the front part of Fig. 13(a'), i.e., the transfer of the mixed voice data signals "V_A + V_B" from the demultiplexer circuit 28 to the multiplexer circuit 22 along the arrow concerned (DMX → MX), a slight delay time inevitably accompanies the related transfer in actuality. However, in the figure, the delay time is disregarded so as to facilitate understanding of the operations. The operations heretofore described relating to the signals "V_A + V_B" also apply to other signals "V_C + V_D" and so on for accumulate the voice data signals.

At the time slot TS_n4, according to the address data nl stored in the memory 25 of Fig. 7 at its address n4, the gist stored in the memory 24 of Fig. 7 at the thus specified address nl is read therefrom as the signals "V_A + V_B" and then supplied to the output buses BSoutl and BS_out2. At the same time, the thus supplied voice dat signals "V_A + V_B" are mixed, at the time slot TS_n5 followed next, with the voice data signal V_C , at the mixing circuit 41, which now appears on the same buses. In this case, the voice data signal V_C is read from the memory 24 of Fig. 7 at its address ℓ, which is specified by the address data ℓ stored in the memory 25 of Fig. 7 at the address n5 corresponding to the time slot TS_n5. As a result, the thus mixed voice data signals at the circuit 41 are provided to the output bus BS_out3 in the form of "VA ⁺ V_B + V_C" and transferred, via the demultiplexer circuit 28 and the multiplexer circuit 22, to the memory 24 again to achieve the related accumulation therein at its address n5 relating to the time slot TS_n5 , as shown by "VA + V_B+ V_C" in Fig. 7.

In a similar way, other mixed voice data signals, such as "V_A + V_B + V_D", are also accumulated and stored in the memory area MA2 for the mixing operation. Thus, the mixing operations, in the frame FR concerned, of the voice data signals from the four subscribers are completed. Soon after this, the mixed voice data signals, for example, "V_A + V_c+ V_D" of Fig. 7, are read according to the address data n9 which is stored in the memory 25 at its address k. The thus read mixed voice data signals "V_A + V_c + V_D" are output, as they are, from the selector (SEL) 44 of the mixing circuit 41, i.e., without passing through the adder 43 of the circuit 41 (refer to the "0" state in Fig. 12(d), at the time slot TS_k in the time slot train TSEX for the mixing operation of the next coming frame FR, to the telephone set 11B to which the time slot TS_k is allotted in advance. Thus, the telephone set 11B receives the mixed voice data signals "V_A + Vc + V_D" as illustrated in Fig. 1. The same applies for the remaining telephone sets 11A, 11C, and 11D. They receive the mixed voice data signals except own side signal, such as shown by "V_B + V_C + V_D", "V_A + V_B + V_D", and "V_A + V_B + V_C in Fig. 1.

The voice data signals with the apostrophes in Figs. 13(b) and 13(c) indicate mixed voice data signals which have been prepared during the preceding frame FR and stored in the memory 24. These mixed voice data signals are also transferred to the respective telephone sets at each time slot TS allotted thereto, respectively. This completes the mixing operation for four subscribers.

The timing controls for both the mixing operation and the line switching operation can be realized by means of the aforesaid timing control unit. Figure 14 illustrates a general construction of the timing control unit. The timing control unit (TG) 29 is mainly comprised of an oscillator OSC, a counter CNTR, and a logic gate circuit LGT. The oscillator OSC generates master clock pulses with which all the operations are synchronized. The counter CNTR operates as a frequency divider to produce relatively low frequency signals for use for various operations for the switching and mixing. The timing control signals for various operations are produced in the logic gate circuit LGT and supplied to the corresponding circuit blocks along the arrows extending outside the unit (TG) 29. The arrows correspond to those in Fig. 4.

Finally, referring again to Fig. 12 (e), it has been stated that nonused time slots 0 are necessarily created by the present invention. Accordingly, the corresponding nonused memory elements ME's are also created in the speech path memory 24 as shown, at the addresses n0, n2, n4, and so on, in Fig. 7. This is due to the provision of mixing pairs MXP's each composed of two successive time slots. The thus created nonused memory elements ME's may be left as they are, but are preferably utilized to accommodate various kinds of digital tones, such as busy tones and dial tones.

As explained above in detail, the switching system of the present invention enables a telecommunication conference with a plurality of subscribers. The number of the subscribers can be freely increased without installing more hardware. Furthermore, it is not required to increase the scale of the mixing trunk along with the increase in the traffic of the telecommunication conferences to be held, which means that the system of the present invention is not influenced by variation of the amount of traffic.