WO2007073234A1

WO2007073234A1 - Method and system for carrying out video-conferences

Info

Publication number: WO2007073234A1
Application number: PCT/RU2006/000225
Authority: WO
Inventors: Grigory Gemfrievich Dmitriev; Alexei Georgievich Sadchikov; Mikhail Yurievich Kochergin; Vladimir Evgenievich Iliin
Original assignee: Grigory Gemfrievich Dmitriev
Priority date: 2005-12-23
Filing date: 2006-04-28
Publication date: 2007-06-28
Also published as: RU2005140355A; RU2321183C2

Abstract

The inventive method and system can be used for carrying out multi-point video/audio conferences using stack transport protocols TCP/IP in broadcasting networks according to a star connection. Said method and system use data flow filtering in such a way that the method consists in substituting a destination address by a multicast group address at a central station during transmission of addresses for each flow which is transmitted from a server and carries information different from another one, in substituting a port by a port on which a videoconference communication device expects to receive a corresponding type of information, wherein the flows carrying similar information are filtered and are not transmitted to a transport network, in substituting a destination address by the address of the videoconference subscriber device located at a given point for flows which are transmissible from the central station and are displayable at this point in a given time during the address transmission on the peripheral points, in filtering the information flows at the peripheral points in such a way that only a specified set of points transmits them to the station, whereas the information flows are filtered and not transmitted at all remaining points.

Description

Method and system for video conferencing.

Technical field

The invention relates to communication technology, and more specifically to a method and system for implementing multi-point video / audio conferences using transport protocols of the TCP / IP stack in the star (point-to-multipoint) scheme in broadcast (broadcast) networks.

State of the art

Currently, there are many different teleconferencing systems (conferencing) that implement the appropriate methods. However, in each of them, an indispensable condition is the use of a video codec to convert the image signal into a digital signal.

For example, in patent EP0353945, class H 04 N 7/15 of 02/07/1990, "Method for establishing multilocal audio / video communication", a specialized system is considered, designed to work with a specific satellite system AT & T SKYNET or with similar systems, providing for direct connection of terminal . devices (codecs) in a satellite modem (with the ability to install an intermediate encryption device).

However, these well-known solutions do not allow multi-point video conferencing using standard video conferencing equipment using TCP / IP stack transport protocols and standard signaling protocols for establishing connections, such as H323, SIP, SCCP and others accepted in IP networks. In addition, they do not allow video conferencing equipment to be located in any of the IP networks served by the satellite station at a considerable distance from the satellite station. At the same time, if there is sufficient bandwidth of the broadcast satellite channel, they cannot simultaneously serve several different video conferences in which various head and terminal devices (codecs) participate.

Known solutions for multi-point broadcasting using standard multicast technology, for example, as described in French patent FR2780229, class H 04 N 7/15 of 12.24.1999, "Multipoint system for the implementation of voice video conferencing." When organizing multi-point video conferencing using multi-point video conferencing servers (hereinafter referred to as MCU - Multiopt Soptrol Upit), which in this patent is referred to as server access, audio / video information is transmitted from terminal devices

(terminals, or codecs) of the conference participants to the MCU ₅ processing information on

MCU and transmission of an integrated stream of audio / video data from the MCU to terminal terminals.

A multicast is understood as - (RPCl 112) a specially allocated range of IP addresses intended for multicast (point-to-multipoint), when using this type of address, several devices can receive the same data packet. In this case, data packets are sent towards those recipient devices that express interest in receiving them. In networks of complex topology, packets with multicast addresses can be duplicated (multiplied) by routers if there are recipients. For routing packets addressed to a multicast group, routing protocols are used that are different from routing protocols for unicast connections when packets are addressed to a specific device and not to a group of devices (point-to-point communication).

At the same time, the integrated data stream going to each terminal is identical in content and differs only in the recipient identifier (in TCP / IP networks, this identifier is the IP address / protocol / port number combination). As a result, in transport networks (for example, satellite ones with a head station or in cable television networks supporting TCP / IP protocol), the bandwidth in the direction from the MCU is extremely inefficient, since the MCU sends several (according to the number of conference participants) identical data streams to the broadcast network for each of the conference participants, while all flows reach each of the terminal terminals, but in reality each terminal accepts only one.

This situation is not very critical for modern high-speed local networks, but it is very significant when using, for example, quite expensive satellite channels. For medium-quality video conferencing, 512K bandwidth per subscriber is required when organizing video conferencing of 10 participants will require a bandwidth in the direction from the MCU of 5 Mb / s, which, moreover, will really be occupied for fairly short periods of time for video conferencing, and the rest of the time idle. Another significant drawback of such systems is the use of multipoint broadcasting using standard multicast technology, which does not allow organizing broadcast conferences using MCUs that do not support multicast, and does not allow the use of another class of devices (H323 or SIP MCU) that do not support multicast in this topology.

Proposed in the patent RU 2240657 Cl, the method and system for the implementation of video conferencing can avoid these drawbacks, however, a number of significant limitations are imposed:

1) All BKC terminals participating in the conference receive an identical stream of video / audio data regardless of the conference mode, which is extremely inconvenient when holding conferences in voice activation mode. In this mode, the server sends to all conference participants an image from the participant speaking at a given time, and to him, in turn, an image of the participant who spoke before him. Thus, all participants in the conference have the impression that they are participating in a dialogue. The limitation is that when using this technology, the participant speaking at a given time will receive his own image;

2) The number of BKC terminals participating in the conference is limited by the number of connections supported by the MCU, which is inefficient in terms of using multicast, as a much larger number of terminals could receive streams of video / audio information without increasing the necessary bandwidth from the MCU to the terminals;

3) The proposed method does not provide the possibility of saving bandwidth on the reverse communication channels, that is, in the direction from the BKC terminals to the MCU, which is very critical for networks based on satellite communication channels. So, for the above conference example for 10 participants, connected at 512 Kb / s each, a satellite resource is required, providing the ability to transfer data at a speed of 512 Kb / s from the MCU to each of the terminals, and 10 * 512 Kb / s in the opposite direction. Thus, the total band used for video conferencing in the mode of simultaneous presence of four subscribers on the screen (in this mode, an identical stream of video information is transmitted to all terminals in the network, in contrast to the voice activation mode) will be 5.5 Mb / s. This gives a gain in comparison with traditional technology, but inefficiently uses the data transmission medium from the point of view that out of 10 reverse video / audio streams, only 4 are really in demand, of which the image transmitted by the server is formed.

Disclosure of a technical solution.

The task to which the claimed invention is directed is to create a method and system for the implementation of video conferencing, which would not contain these disadvantages.

In solving this problem, a technical result is achieved, consisting in a significant reduction in the required total bandwidth of data channels from remote video conferencing terminals (BKC) to a multipoint video conferencing server (MCU) by using filtering mechanisms for data streams transmitting video and voice information;

Another technical result achieved by the claimed invention is the ability to serve a much larger (almost unlimited) number of BKC terminals working to receive video / audio information from the server ("passive" mode) than the multi-point video conferencing server allows, which determines only the number of terminals working in the "active" mode (for receiving and transmitting information), using mechanisms for modifying signal information. Another technical result achieved by the claimed invention is the ability to change the set of "active" / "active" participants not only during the organization of the conference, but also during its holding.

These technical results are achieved by the proposed method and system for video conferencing according to the independent claims and dependent subordinates in all areas of the alternatives contained in the claims.

The basic principle used in the proposed technology is the forced filtering of all MCUs sent to the Broadcast network of audio / video data streams to subscriber devices (codecs), except for information other than the others, and address translation on the receiving side so that in the future each subscriber the codec perceived this data stream as intended for it. The return channels used for data transmission from subscriber codecs to the MCU can be organized in any way, this is not essential for the proposed technology, since the transmitted information is unique for each subscriber codec.

According to the proposed method for video conferencing between at least three devices:

- as the transport protocol using the protocol stack TCP / IP;

- as a transport network, use a network with support for multicast messaging - multicast;

- at the central node for processing and generating audio / video streams, a multipoint conference organization server is used that supports standard signaling protocols for establishing connections when exchanging audio-video information in IP networks - H.323, SIP, SCCP and others, in which signaling and audio / video streams are transmitted with a different combination of ip-address / port / protocol;

- at peripheral points, video conferencing terminal devices use devices that support standard signaling protocols for establishing connections when transmitting audio / video streams in IP networks - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are transmitted with different IP address / port / protocol combination;

- at the central node translate addresses for streams of audio / video information coming from the video conferencing server, so that for each stream, that carries information other than the other, the destination address is changed to the multicast group address, and the port to the one on which the video conferencing device expects to receive the corresponding type of audio / video information, while streams with similar information are filtered and not transmitted to the transport network;

- at the peripheral points, the addresses for the streams of audio / video information coming from the central node are translated, so that for the streams that should be displayed at this point at a given time, the destination address is replaced with the address of the video conferencing device installed at this point;

- at peripheral points, the filtering of streams of audio / video information is performed in such a way that only a given set of points transfers them to the side of the node, for all other streams of audio / video information is filtered and not transmitted to the transport network.

At peripheral points, for each stream that carries information other than the others, the destination address can be changed to the address of the multicast group, and the port to the one on which the video conferencing server expects to receive the corresponding type of audio / video information.

Streams of audio / video information are delivered to a video conferencing device connected to a subscriber device using standard protocols - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are addressed with a different combination of ip-address / port / protocol.

The number of terminals connected to the conference may be greater than the number of occupied ports of the video conferencing server.

To connect video conferencing devices, a signal exchange protocol may be used that is different from that used by the video conferencing server.

The set of video conferencing terminal devices can be changed during the conference.

During the conference, at various points in time, various streams of audio / video information can be broadcast to the address of the video conferencing device. The maximum number of videoconferencing devices participating in one conference does not depend on the number of clients supported by the videoconferencing server.

A transport network can be any network that supports broadcasting in accordance with the TCP / IP protocol stack, for example, a satellite network, a data network based on a cable television network, or an urban broadband access network.

Also proposed is a system for video conferencing between at least three video conferencing devices, in which:

- as the transport protocol using the protocol stack TCP / IP;

- at the central node they use a server for organizing multipoint conferences for processing and generating audio / video streams, supporting standard signaling protocols for establishing connections when exchanging audio-video information in IP networks - H.323, SIP, SCCP and others, in which signaling and audio / video streams are transmitted with a different combination of ip-address / port / protocol;

- translation of addresses on the central node for streams of audio / video information coming from the video conferencing server is performed in such a way that for each stream carrying information other than the other, the destination address is replaced with the address of the multicast group, and the port with the one on which the video conferencing device expects to receive the corresponding type of audio / video information; while streams with similar information are filtered and not transmitted to the transport network;

- translation of addresses at peripheral points for streams of audio / video information coming from the central node is made in such a way that for streams that should be displayed at this point at a given time, destination address is replaced with the address of the video conferencing device installed at this point;

- filtering streams of audio / video information at peripheral points is performed in such a way that only a given set of points can transmit them to the node side, and on all other streams of audio / video information is filtered and not transmitted to the transport network.

For each stream carrying information other than the other, the destination address can be changed to the multicast group address, and the port to the one on which the video conferencing server expects to receive the corresponding type of audio / video information.

The streams of audio / video information are delivered to a video conferencing device connected to a subscriber device using standard protocols - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are addressed with a different combination of ip-address / port / protocol.

The system may contain a unit modifying the signaling information in such a way that the number of terminals connected to the conference may be greater than the number of occupied ports of the video conferencing server.

During the conference at the peripheral points, the rules for broadcasting multicast audio / video streams can be changed in such a way that for those streams that should be displayed at this point at a given time, the address of the multicast group is replaced with the address of the video conferencing device installed at this point;

The maximum number of videoconferencing devices participating in one conference does not depend on the number of clients supported by the videoconferencing server.

Any network that supports broadcasting in accordance with the TCP / IP protocol stack, for example, can serve as a transport network. a satellite network, a cable television network, or an urban broadband access network.

Elements of a system that implements the proposed method and system are:

1) a mechanism for establishing a connection with BKC remote terminals with a special server that converts and multiplies signal information flows (SIP, H323 or others) in such a way that the connections are established from the MCU “name”;

2) a mechanism for ensuring the terminal’s operation in the “passive” mode, which performs selective filtering of the transmitted data packets from the terminal to the MCU based on the address, protocol and port, and selective, based on the address, protocol and port, translation of the addresses of the data packets transmitted from the MCU to the peripheral devices

3) a mechanism for controlling devices that implement the functions of routing, filtering and address translation at peripheral points, allowing in real-time mode or according to a predetermined program (schedule) to enable / disable streams of audio / video data from remote BKC terminals, that is, transmit streams to the MCU audio / video data only from “active” terminals, which can significantly save bandwidth;

4) the control mechanism of the BKC and MCU terminals, providing the operator with access to their user functions;

5) the mechanism for connecting BKC terminals to video selection devices according to the protocols adopted for BKC terminals (SIP, H323, etc.) with their subsequent conversion to the protocol used to connect to a special server, except when these protocols coincide.

These mechanisms can be implemented on separate devices or integrated into video conferencing devices. For simplicity, we will consider the case when they are implemented on separate devices, for simplicity called video selection devices.

A distinctive feature of the proposed method and system is the use of a fixed total bandwidth that depends only on the type of conference and does not depend on the total number of conference participants. So, for the above conference example for 10 participants connected to 512 Kb / s, in the mode of simultaneous presence of 4 participants, the necessary bandwidth in the direction from the MCU to the terminals will be 512 Kbit / s, and in the direction from the terminals to the MCU - 4 * 512 Kbit / s, that is, the total bandwidth required for the conference will drop to 2.5 Mbps. At the same time, due to signaling processing by the proposed method, with an increase in the number of conference participants to 20, the requirements for network bandwidth will not change, and the number of ports occupied on the MCU will not increase.

The main requirement for network infrastructure for the effective use of the method is the presence of a broadcast “direct” channel from one (any) of the points participating in the video conference, conventionally called central to the other participants in the video conference. The essence of using a broadcast channel is known from the literature and patents. At the same time, audio-video streams go to the central point only from those points that other conference participants should see (that is, from points that are in the "active" mode). Any communication channels using the TCP / IP protocol stack as a transport protocol can be used as reverse. Multicast support on these channels is optional.

From the video selection devices, the system management server manages the devices in such a way that they all receive information transmitted via the satellite channel, but the transmission is only from “currently active” users. To organize standard video conferencing management functions between “active” participants (such as voice activation or simultaneous presence mode), a standard video conferencing server (MCU) is used that supports the required number of active participants. Moreover, the total number of participants can be significantly larger than the MCU itself supports.

A brief description of the drawings.

The general distribution scheme of video / audio streams in a videoconferencing network constructed by the proposed method is shown in FIG. 1 (for voice activation mode) and in FIG. 2 (for simultaneous presence mode). Variant of the technical solution

In the central point »1 (which may not be the central office, is selected from the point of view of the network topology) a video selection device is installed

2, a multi-point video conferencing server 3 (MCU). Optionally in

A “central point” can also be set to the video codec (s) connected directly to the MCU.

At the peripheral points, video selection devices 2 and video conferencing terminals are installed, which can be in “active” 4 or “active” 5 modes.

At the same time, several information streams are used to transmit video / audio information, depending on the conference mode. In voice activation mode (Fig. 1), a multicast stream from the MCU to all passive terminals 6 is used, unique streams from the MCU to active terminals / from active terminals to the MCU 7, 8. In the simultaneous presence mode, a multicast stream from the MCU to all terminals is used 6 and reverse flows from active terminals to MCU 7, 8.

The video selection device consists of the following functional blocks:

- an operator interface that allows you to set conference parameters, select active participants, switch active / inactive, etc .;

- MCU control unit transmitting conference parameters to the MCU in accordance with the parameters entered from the operator interface;

- control unit for remote terminals BKC;

- a control unit for remote video selection devices that convert the resulting broadcast audio / video stream to a stream intended for a given video conferencing terminal, as well as enable / disable the transmission of the audio / video data stream to a central point depending on the selected participant state (active / inactive); and also, if necessary, the translation of the control protocol in an acceptable for the connected terminal (SIP, H323, etc.)

- control unit for a special server;

- routing block;

- filtration unit;

- address translation unit; - diagnostic unit;

- a special server unit that connects remote points from the "name" of the MCU, that is, connect in the same way as if they were connected directly to the MCU.

In addition, the video selection device control system provides other auxiliary functions and functions for the on-line diagnostics of the entire device network.

The management of the network of video selection devices is organized in such a way that the operator can be located at any convenient point that has a connection with the "central" point.

Thanks to the module for converting protocols adopted for controlling BKC terminals (SIP, H323, etc.) into a control protocol used between video selection devices, it is possible to connect any terminal supporting standard control protocols to video selection devices. This feature allows you to organize video conferences with the participation of terminals using various control protocols (for example, H323 and SIP).

Thus, the proposed method and system for organizing multipoint video conferencing has the potential to significantly save the total bandwidth (for example, with a total number of participants 10 and the simultaneous presence of 4 active participants, the total required bandwidth will be 5 * M (where M is the transmission rate for one audio / video stream), 4 streams from active participants and one broadcast stream to all participants, while with a standard configuration the total band would be 20 * M (one stream each participant from the MCU and one from each participant to the MCU.) That is, even for a low-quality video conferencing speed of 256K, the savings in the total bandwidth will be 20 * 256-5 * 256 = 3.384 Mb / s, using the total band of 1.28 Mb /from.

The proposed method and system allows organizing multipoint video conferences both in the activation mode by voice (voice activation) and in the mode of simultaneous presence (voice protection). In the first mode, “Active” participants are predefined (from two or more, their number is less than the total number of conference participants, but does not exceed the supported MCU); selection of the terminal from which the image is transmitted to all other participants, based on who is currently speaking. During the conference, the composition of active / passive participants can be changed by the operator. In the second case, the video image from predefined participants is transmitted simultaneously in a split screen mode (the total number is less than the total number of conference participants, but does not exceed the supported MCU), voice streams from the active participants are mixed. During the conference, the composition of active / passive participants can also be changed by the operator.

Typical algorithms for preparing and conducting a videoconferencing session using the system in both modes are shown in FIG.

In the activation mode by voice, the speaker should not see his own image, but the previous speaker, therefore, each active subscriber will be sent a multimedia stream in which, depending on the terminal status (voice / silence), the speaker’s audio / video image is transmitted moment or previous. Since there are passive clients who always need to transfer only the speaker at the moment, a third stream is used for broadcast broadcasting, which cannot be addressed to the speakers. Wherein:

- on all peripheral nodes corresponding to passive BKC terminals, standard multicast -> unicast translation tables for audio and video streams (hereinafter referred to as “RTP streams”) distributed from the center in the multicast group are activated, with the address of the corresponding BKC terminal being substituted instead of the multicast group;

- on all peripheral nodes corresponding to passive BKC terminals, standard filters are activated that prohibit the transmission of audio and video streams from the BKC terminal;

- on the central node, a standard table for translating outgoing RTP streams to multicast is activated for broadcasting audio / video to all passive BKC terminals;

- a command is given on the MCU to connect participants to the conference;

- the alarm management module connects participants, while due to signaling processing the number of participants connected to the conference can be more than the number of ports used on the MCU; - based on the port data for initially active BKC terminals, port translation tables for incoming RTP streams are generated and activated on the central node so that the MCU receives these streams on standard ports;

Thus, all participants except the speaker see and hear the speaker at the moment, and the speaker sees the previous speaker.

If it is necessary to change one or more active users, the operator from the console gives the appropriate command, as a result of which execution on the peripheral nodes:

- turning into a passive state, the used translation tables are deactivated, standard filters that prohibit the transfer of audio and video streams from the BKC terminal are activated, and standard multicast -> unicast translation tables for RTP streams distributed from the center to the multicast group are activated;

- transitioning to an active state, filters for reverse RTP streams are removed and translation tables are set in accordance with the addresses and ports of participants switching to passive mode.

In the simultaneous presence mode, both active and passive subscribers should see (windows with images of active participants) and hear (mixed voice) the same thing. Moreover, for the formation of a broadcast multicast channel for transmission to all participants of RTP streams, streams intended for one of the clients are used. The remaining RTP streams transmitted from the center are not used. Wherein:

- on all peripheral nodes BKC terminals are activated standard multicast translation tables -> unicast for audio and video streams distributed from the center in the multicast group, with the substitution of the address of the corresponding BKC terminal instead of the multicast group;

- on all peripheral nodes corresponding to passive BKC terminals, standard filters are activated that prohibit the transfer of audio and video streams from the BKC terminal;

- on the central node, a standard table for translating outgoing RTP streams to multicast is activated for broadcasting audio / video to all passive BKC terminals; - a command is given on the MCU (codec with built-in MCU) to connect participants to the conference;

- the alarm management module connects participants, while due to signaling processing the number of participants connected to the conference can be more than the number of ports used on the MCU;

- based on the port data for initially active BKC terminals, port translation tables for incoming RTP streams are generated and activated on the central node so that the MCU receives these streams on standard ports;

In case of successful completion of these actions, the conference begins; in case of malfunctions, an error message is issued to the operator.

Thus, all participants receive the same RTP streams, and accordingly see and hear the same thing, i.e. active participants in the windows and mixed voice.

- turning into a passive state, the used translation tables for RTP streams towards the center are deactivated and standard filters are activated that prohibit the transfer of audio and video streams from the BKC terminal;

- transitioning to an active state, filters for reverse RTP streams are removed and translation tables for reverse RTP streams are set, in accordance with the addresses and ports used by BKC terminals that are in passive mode.

Industrial applicability

The technical solution reduces the required total bandwidth of data channels from remote video conferencing terminals (BKC) to a multipoint video conferencing organization (MCU) server by using filtering mechanisms for data streams transmitting video and voice information.

Claims

Claim.

1. A method of video conferencing between at least three video conferencing devices, in which:

- as the transport protocol using the protocol stack TCP / IP;

- at peripheral points, video conferencing terminal devices use devices that support standard signaling protocols for establishing connections when transmitting audio / video streams in IP networks - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are transmitted with different IP address / port / protocol combination, characterized in that:

- on the central node, the addresses for the streams of audio / video information coming from the video conferencing server are translated, so that for each stream carrying information other than the others, the destination address is replaced with the address of the multicast group, and the port with the one on which the video conferencing device expects to receive the corresponding type of audio / video information, while streams with similar information are filtered and not transmitted to the transport network;

- at peripheral points, the filtering of streams of audio / video information is performed in such a way that only a given set of points transfers them to the side node, on all other streams of audio / video information is filtered and not transmitted to the transport network.

2. The method according to claim 1, characterized in that at the peripheral points for each stream carrying information other than the other, the destination address is changed to the address of the multicast group, and the port to the one on which the video conferencing server expects to receive the corresponding type of audio / video information.

3. The method according to any one of paragraphs. 1, 2, characterized in that the streams of audio / video information are delivered to a video conferencing device connected to a subscriber device using standard protocols - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are addressed with a different combination of ip address / port / protocol.

4. The method according to any one of paragraphs. I ₅ 2, characterized in that the number of terminals connected to the conference may be greater than the number of occupied ports of the video conferencing server.

5. The method according to any one of paragraphs. 1, 2, characterized in that for connecting video conferencing devices use a signal exchange protocol that is different from that used by the video conferencing server.

6. The method according to any one of paragraphs. 1, 2, characterized in that the set of video conferencing terminal devices is changed during the conference.

7. The method according to any one of paragraphs. 1, 2, characterized in that during the conference at different points in time to the address of the video conferencing device broadcast various streams of audio / video information.

8. The method according to any one of paragraphs. 1, 2, characterized in that the maximum number of video conferencing devices participating in one conference does not depend on the number of clients supported by the video conferencing server.

9. The method according to any one of paragraphs. I ₅ 2, characterized in that the transport network is any network that supports broadcasting in accordance with the TCP / IP protocol stack, for example, a satellite network, a data network based on a cable television network or an urban broadband access network.

10. A system for video conferencing between at least three video conferencing devices, in which:

- as the transport protocol using the protocol stack TSNYAR; - as a transport network, use a network with support for multicast messaging - multicast;

- at peripheral points, video conferencing terminal devices use devices that support standard signaling protocols for establishing connections when transmitting audio / video streams in IP networks - H.323, SIP, SCCP and others, in which the signaling and audio / video streams are transmitted with different IP address / port / protocol combination; characterized in that in it:

- translation of addresses at peripheral points for streams of audio / video information coming from the central node is performed in such a way that for streams that should be displayed at this point at a given time, the destination address is replaced with the address of the video conferencing device installed at this point ;;

11. The system of claim 10, characterized in that for each stream carrying information other than the other, the destination address is changed to the address of the multicast group, and the port to the one on which the video conferencing server expects to receive the corresponding type of audio / video information.

12. The system according to any one of paragraphs. 10, 11, characterized in that the streams of audio / video information are delivered to a video conferencing device connected to a subscriber device using standard protocols - H.323, SIP ₅ SCCP and others, in which the signaling and audio / video streams are addressed with a different combination of ip address / port / protocol.

13. The system according to any one of paragraphs.10, 11, characterized in that it contains a block modifying the signal information so that the number of terminals connected to the conference may be greater than the number of occupied ports of the video conferencing server.

14. The system according to any one of paragraphs.10, 11, characterized in that for the connection of video conferencing devices use a signal exchange protocol different from that used by the video conferencing server.

15. The system according to any one of paragraphs.10, 11, characterized in that the set of video conferencing terminal devices is changed during the conference.

16. The system according to any one of claims 10, 11, characterized in that during the conference at the peripheral points, the rules for broadcasting multicast audio / video streams are changed in such a way that for those streams that should be displayed at this point at a given time , the address of the multicast group is replaced with the address of the video conferencing device installed at this point;

17. The system according to any one of items 10, 11, characterized in that the maximum number of video conferencing devices participating in one conference does not depend on the number of clients supported by the video conferencing server.

18. A system according to any one of claims 10, 11, characterized in that any network supporting broadcasting in accordance with the TCP / IP protocol stack, for example, a satellite network, a data transmission network based on a cable television network, acts as a transport network or urban broadband access.