JP2003258776A - Communication device, communication system and video conference system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication device and a communication system in which a communication quality can be guaranteed by packet encoding processing, and to provide a video conference system using the same. <P>SOLUTION: The communication device has a redundant packet generating means for generating a redundant packet by a prescribed encoding system in response to an information packet to be transmitted and a packet recovery means for recovering a lost packet on the basis of the redundant packet and the information packet received by a recovery method corresponding to the encoding system when the loss of the packet is detected in a received packet. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication device, a communication system and a video conference system using the Internet.

[0002]

2. Description of the Related Art A so-called Internet video conference system can be realized by transmitting and receiving image (moving image) and audio information between a plurality of terminals in real time using the Internet. Since the Internet video conference system transmits information using the Internet, it is not necessary to provide a dedicated line, and the video conference system can be constructed at low cost. Up to now, various Internet video conference systems have been proposed and some have been put into practical use. For example, NetMeeting of Microsoft Corporation is known as a one-to-one conference system. It is connected to the internet 2
A video conference between two parties is realized by transmitting audio and video information between one client terminal.

Further, as a video conference system between a plurality of points, a system centrally managed by an MCU (Multipoint Control Unit), for example, "MCS2000 series" developed by NTT-ME, or "View Station developed by Polycom". Series ”etc. are known. All of these systems are JT-H32
3: Packet-based multimedia communication system (I
TU-T Recommendation H. 323). In addition, as a one-to-many communication system, a system such as remote lecture between remote campuses, for example, a system of Stanford University in the United States is known.

[0004]

Data transmission on the Internet is so-called packet communication in which information data is divided into communication units called packets and transmitted. Depending on the state of the communication path, a part of the packet sent by the sender may be lost, resulting in packet loss that cannot reach the receiver. The communication protocol that normally manages the transmission of information data over the Internet is
By detecting lost packets and requesting the sender to resend them, the lost packets are recovered and the communication quality is maintained.

However, the above-mentioned conventional Internet video conference system needs to transmit a large amount of audio and moving image information in real time, and therefore UDP (User Datag
High-speed communication can be realized by using a communication protocol called ramProtocol), but effective means for compensating for packet loss that may occur in the communication path has not been taken. Therefore, the reliability of communication is not guaranteed, and a part of the screen may be disturbed or the voice may be interrupted during the use of the system. Furthermore, since the MCU server manages the entire system of video conferencing, traffic concentrates on this part, packet switching is not performed smoothly, and
The disadvantage is that the failure of the MCU leads to an outage of the entire system.

The present invention has been made in view of the above circumstances, and an object thereof is to apply error correction coding to a transmission packet and recover a packet lost during data transfer on the receiving side to improve communication quality. (EN) Provided are a communication device, a communication system, and a video conferencing system capable of improving the reliability of the entire system by improving and improving the reliability of the entire system by decentralizing and managing the entire system to mitigate traffic concentration and arbitrating communication between clients. Especially.

[0007]

In order to achieve the above object, the communication apparatus of the present invention comprises a redundant packet generating means for generating a redundant packet by a predetermined encoding method in accordance with an information packet to be transmitted, and a receiving means. A packet recovery means for recovering the lost packet based on the redundant packet and the information packet received by the recovery method corresponding to the encoding method when the packet loss is detected in the packet.

The communication system of the present invention further comprises: redundant packet generating means for generating a redundant packet by a predetermined coding method in accordance with an information packet to be transmitted; The information packet and the redundant packet are transmitted through a communication network using a communication device having a packet recovery means for recovering the lost packet based on the redundant packet and the information packet by a recovery method corresponding to an encoding method. To do.

Further, in the present invention, it is preferable that the redundant packet generating means controls the coding rate according to the occurrence status of the packet loss to generate the redundant packet. Further, the redundant packet is generated by changing the generation matrix or constraint length used in the encoding process according to the occurrence status of the packet loss.

Further, in the present invention, it is preferable that the redundant packet generating means sets different coding rates for different types of information data to be transmitted and performs a coding process.

Further, in the present invention, it is preferable that the redundant packet generating means generates a redundant packet for the variable length packet in accordance with the maximum packet length of the information packets to be encoded.

Further, in the communication system of the present invention, the redundant packet generation means embeds predetermined data in the excess bits of the information packet shorter than the maximum packet length to make the maximum packet length.

Further, in the present invention, preferably, there is provided a transmitting means for transmitting the information packet and the redundant packet through different communication channels.

Also, the video conference system of the present invention generates an information packet based on predetermined information data, transmits the information packet through the communication network, and responds to the information packet transmitted through the communication network. A terminal for reproducing information data, a relay means for connecting to a plurality of the terminals via a communication network, transmitting the information packet transmitted from the plurality of terminals to a communication control means, and the communication means, A destination list indicating a plurality of destinations is added to a predetermined area of the packet transmitted from the relay unit, and the packet is transmitted to the relay unit.

Also, in the video conference system of the present invention,
Preferably, the relay means transmits the packet to the designated terminal when the terminal connected to itself is designated by the destination list in the received packet.

Further, in the video conference system of the present invention,
Preferably, the relay means transmits the packet to the designated terminal when the terminal connected to itself is designated by the destination list in the received packet.

In the video conference system of the present invention,
Preferably, the relay means transmits the packet to another relay means when the terminal connected to itself is not designated by the destination list in the received packet.

Also, in the video conference system of the present invention,
Preferably, at least two communication control means are provided, and when one of the communication control means fails, the other communication control means performs the communication control.

In the video conference system of the present invention,
Preferably, the terminal generates a redundant packet based on the information packet in accordance with a predetermined coding method, and transmits the redundant packet to the relay unit together with the information packet.

Further, in the video conference system of the present invention, preferably, the communication control means generates a redundant packet in accordance with a predetermined encoding method based on the information packet sent from the relay means. , And transmits to the relay means together with the information packet.

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Congestion occurs on the Internet when packets exceeding the processing capability of a router flow in. At this time, the packet overflowing from the receiving buffer of the router is not processed and is discarded, which causes a packet loss. Further, when an error occurs in a packet during transmission when the communication quality of the transmission line is degraded, the error may be detected in the data link layer and the packet may be discarded. Therefore, it is important to efficiently recover lost packets in order to guarantee the reliability of the information to be transmitted.

TCP (Transmission Control Protocol) which is usually used for packet communication on the Internet
In (), the lost packet is recovered by an automatic repeat method by ARQ (Automatic Repeat reQuest) or the like to improve reliability. However, although TCP guarantees reliable data transmission, when a packet loss occurs, it requests the transmitting side to retransmit the lost packet, resulting in a decrease in the data transfer rate. In particular, when the quality of the communication path is low and packet loss frequently occurs, the data transfer rate is significantly reduced.
Therefore, UDP is used instead of TCP for an application that transmits a large amount of information data at high speed.

However, as described above, when UDP is used, packet loss is not considered at all, and communication quality is not guaranteed. The recovery of lost packets is left to the upper layer applications. For this reason, it is necessary for the application to detect the packet loss that occurred during transmission and recover the lost packet as needed. Therefore, in the communication system of the present invention and the video conference system using the same, the transmitting side generates an error correction redundant packet by convolutional coding, and the receiving side performs decoding processing when packet loss is detected, resulting in loss. Recover the packet that you made. As a result, high-speed data communication is realized while maintaining communication quality, and a video conference system using the Internet is realized.

Each embodiment of the communication system of the present invention and the video conference system using the communication system will be described below with reference to the drawings.

First Embodiment FIG. 1 is a block diagram showing a first embodiment of a communication system according to the present invention. As shown in the figure, the communication system of this embodiment is composed of a plurality of data communication devices 100-1 and 100-2 connected to a communication network 110. The data communication device is incorporated in, for example, a server or a communication terminal, and functions as one component of the server or the terminal. The communication network 110 is a LAN (Lo
cal Area Network) This means any communication network such as WAN (Wide Area Network), generally the Internet. Further, although two data communication devices are shown in FIG. 1, a large number may be used.

As shown in FIG. 1, the data communication device comprises a packet coding device 10, a packet reproducing device 20 and a network interface 30. The packet encoding device 10 includes a packet generation unit 12 and a redundant packet generation unit 14, and the packet reproduction device 20 includes a data reproduction unit 22 and a lost packet recovery unit 24. In addition, in the communication system of the present embodiment, the data communication device is not limited to the above-described configuration, and other modified examples are possible.
For example, in a special case where only the transmission is performed, as an example, in a device such as a net broadcasting station that distributes image or audio information via the Internet, the packet encoding device 1
0 and network interface 30 are all you need,
Further, when only receiving is performed, the packet reproducing device 20 and the network interface 30 are sufficient.

The respective constituent parts of the packet encoding device 10 and the packet reproducing device 20 will be described below. The packet encoding device 10 performs packet encoding processing based on the information packet. In the packet encoding device 10, the packet generator 12 divides the information data D _S into a predetermined length to generate an information packet. A header made up of additional information necessary for each information packet is added. The content of the header can include information predetermined by the communication system in addition to the content defined by the communication protocol. For example, when communication is performed using UDP, the header includes an IP (Internet Protocol) header, and a source port, a destination port, a packet length, and a checksum defined by UDP.
It also includes a sequence number for correctly reproducing the original information data D _S on the packet receiving side. The information packet thus generated is sent to the network interface 30 together with the redundant packet generated by the redundant packet generation unit 14 and transmitted via the communication network 110.

The redundant packet generator 14 generates a redundant packet according to the information packet. As mentioned above,
In the communication system of this embodiment, redundant packets are generated by convolutional coding. Convolutional coding is a method of generating a redundant code by a predetermined logical operation based on original data using a predetermined generator matrix. The redundant packet generation unit 14 of the present exemplary embodiment generates a redundant packet based on the information packet by, for example, an exclusive OR operation for each bit in a generation matrix determined in advance by the communication system.

FIG. 2 is a conceptual diagram showing the generation of redundant packets. As illustrated, here, for example, k (k ≧ 1, k is an integer, the same applies to the following symbols n and m) information packets P1 and P according to the information data.
2, ..., Pk are generated. Then, based on the k information packets, nk redundant packets are generated using a predetermined generation matrix. That is, a total of n packets are generated based on k information packets. A sequence number is assigned to each of these packets, and the sequence number is stored in the header together with other additional information. The n packets thus generated are sequentially transmitted by the network interface 30.

Next, the packet reproducing device 20 will be described. In the packet reproduction device 20, the lost packet recovery unit 24 detects whether or not there is a loss in the information packet sent by the transmission side, and if there is a lost packet, based on the redundant packet and other information packets, Recover lost packets.

Then, the data reproducing section 22 reproduces the original information data based on the packet recovered by the lost packet recovering section 24 from the received information packet.

Next, generation of redundant packets in the redundant packet generator 14 will be described in more detail. Figure 3
FIG. 6 is a diagram showing a redundant packet generation process based on an information packet. The redundant packet is generated by a coding process using a predetermined generation matrix in units of packet groups each including a plurality of information packets.

Here, an example in which n−k redundant packets are generated for every n information packets will be described. As shown in FIG. 3, n−k redundant packets are generated for k information packets. Then, the generated n packets are sequentially transmitted. That is, in this case, the coding rate in packets, that is, the number of information packets / the number of packets obtained by the coding process is k / n. The higher the coding rate, the smaller the number of redundant packets and the higher the efficiency of packet transmission, but the ability of the receiving side to recover a lost information packet decreases. On the other hand, when the coding rate decreases, more redundant packets are generated for a predetermined number of information packets, and the transmission efficiency decreases, but the receiving side improves the ability to recover lost information packets.

Next, a generator matrix for generating redundant packets will be described. The generator matrix is a matrix G having k rows and m + 1 columns. Each element of the generator matrix is either 1 or 0. When generating a redundant packet, the information packet corresponding to element 1 of the generator matrix is used for the logical operation, while the information packet corresponding to element 0 of the generator matrix is not used for the logical operation. The number of elements in the generator matrix is called the constraint length for generating redundant packets. For example, when using a generator matrix of k rows and m + 1 columns, the constraint length is k (m + 1). As shown in FIG. 3, for example, the redundant packet PK _C
Are generated based on k (m + 1) information packets immediately before. That is, the redundant packet PK _C is related only to k (m + 1) information packets determined by the constraint length, and is not related to other packets.

The convolutional coding process described above is
It is called a convolution process using a (n, k, m) convolutional code. That is, here, n represents the number of packets obtained by the encoding process, k represents the number of original information packets, and m represents the number of columns of the generator matrix (the generator matrix has m + 1 columns). Also, m
And k determine the constraint length k (m + 1).

Here, as an example, a convolutional coding process using a (3,2,1) convolutional code will be described.
Generation of redundant packets by convolutional encoding and recovery of lost packets will be described. The generator matrix used in this convolutional encoding process is given by the following equation.

[0037]

[Equation 1]

In this case, the constraint length is 2 (1 + 1) = 4. That is, the redundant packet is generated based on the immediately preceding four information packets.

FIG. 4 shows an example of an information packet and a redundant packet generated based on the information packet. As shown in FIG. 4, one redundant packet is generated for every two information packets. The redundant packet is generated by an exclusive OR operation for each bit of the information packet using the generator matrix G shown in Expression (1) based on the immediately preceding four information packets. Note that, in FIG. 4, the information packet and the redundant packet include a header portion including only a sequence number and
It is simply shown as a packet having only bits of valid data, which is different from the actual packet.

In FIG. 4, PK1, PK2, PK4 and PK5 indicate information packets, and PK3 and PK6 indicate redundant packets. The redundant packet PK3 is
Since it is generated based on the packets PK1 and PK2 and the information packet (not shown) preceding the packets PK1 and PK2, the contents thereof are not particularly specified. The redundant packet PK6 is generated based on the information packets PK1, PK2, PK4 and PK5. When the generator matrix G shown in Expression (1) is used, the information packets used to generate the redundant packet PK6 are only the information packets PK1, PK4, and PK5. Information packet PK2 corresponding to element 0 of the generator matrix
Is not used.

As shown in FIG. 4, each bit of the information packet PK6 is calculated as an exclusive OR of the respective bits of the information packets PK1, PK4 and PK5. The redundant packet thus generated is transmitted together with the information packet via the communication network.

Next, with reference to FIG. 5, a description will be given of a process of recovering a packet lost on the transmission path on the receiving side based on the above-mentioned information packet and the redundant packet generated according to the information packet.

FIG. 5 is a schematic diagram for explaining the lost packet recovery process. As illustrated, the receiving side sequentially receives the packets encoded by the transmitting side. Here, for example, among the six packets shown in FIG. 4, it is assumed that the information packet PK5 is lost during transmission and is not transmitted to the receiving side. On the receiving side, based on the sequence number in the header of each packet received,
The loss of the information packet PK5 can be detected.

Then, the lost packet recovery unit 24 on the receiving side
At, the lost information packet PK5 is recovered on the basis of the received other information packet and the redundant packet. It is known on the receiving side that among the received packets, the packets PK3 and PK6 with sequence numbers 3 and 6 are redundant packets, and the other packets are information packets. Further, the redundant packet PK6 is the information packet P.
It is also known in advance that it was generated using the generator matrix G shown in Expression (1) based on K1, PK4 and the lost information packet 5. Therefore, the lost packet recovery unit 24
Are information packets PK1, PK4 and redundant packet PK
Based on 6, the information packet PK5 can be recovered.

As shown in FIG. 5, the redundant packet PK6
Is calculated by the bitwise exclusive OR operation of the information packets PK1, PK4 and PK5, so that each bit of the information packet PK5 lost by the logic operation can be calculated. In consideration of the reversibility of the exclusive OR operation, the information packet PK5 is equivalent to the information packets PK1 and PK.
4 and redundant packet PK6 can be easily obtained by an exclusive OR operation for each bit.

As described above, in the communication system of this embodiment, in the packet coding device 10 on the transmission side, a redundant packet is generated by convolutional coding based on the information packet and the generator matrix, and both the information packet and the redundant packet are generated. Sent over a communications network. Then, in the packet reproducing device 20 on the receiving side,
When the loss of the information packet is detected, the lost information packet can be recovered based on the other information packet and the redundant packet and the known generation matrix, so that the reliability of the communication system can be improved.

Second Embodiment Next, a second embodiment of the communication system of the present invention for adaptively controlling the coding rate, the constraint length, the generation matrix and the like according to the communication state.
Will be described. In a communication system that performs packet coding processing and lost packet recovery processing,
As mentioned above, the communication efficiency changes depending on the coding rate,
The ability to recover loss also changes. For example, if the coding rate is set to a large value, the communication efficiency can be maintained high, but the ability to recover lost packets will be reduced. On the contrary, when the coding rate is set to be small, the communication efficiency is reduced, but the lost packet recovery capability is improved. In other words, in a communication system, there is a trade-off relationship between communication efficiency and loss recovery capability, and by optimally controlling the coding rate according to the status of the communication path, maximum communication efficiency is achieved while maintaining communication quality. it can.

FIG. 6 is a block diagram showing the configuration of the communication system according to the second embodiment of the present invention. In the data communication device of the present embodiment, the coding rate,
Optimal control of constraint length or generator matrix. As shown in FIG. 6, in the communication system of this embodiment, a data communication device 1 that communicates with each other via a communication network 110.
00-3 and 100-4, and communication devices 100-3 and 1
Between 00-4 and 00-4, in addition to normal packet communication, information transmission for optimizing the coding rate is also performed.

In this embodiment, the data communication device 10
0-3 is composed of an encoder 40, a changing unit 50, a monitor / notification unit 60, a decoder 70, and an application unit 80. The data communication device 100-4 also has the same configuration. The encoder 40 is, for example, as shown in FIG.
The data communication device 100 according to the first embodiment of the present invention shown in FIG.
The packet generator 12 and the redundant packet generator 14 in -1 are provided. Also, the decoder 70
Includes both the data reproducing unit 22 and the lost packet recovery unit 24 in the data communication device 100-1 of the first embodiment. However, the encoder 40 of the present embodiment generates a redundant packet based on the coding rate, the constraint length, or the generation matrix instructed by the changing unit 50.

The changing unit 50, based on the notification from the monitor / notification unit 60 of the data communication device 100-4 of the communication partner,
The encoder 40 is instructed of the coding rate, constraint length, or generation matrix for generating redundant packets. The instruction of the generation matrix specifies, for example, a predetermined one of a plurality of generation matrices prepared in advance by the communication system using an identification number or the like. Here, it is assumed that an identification number is previously assigned to each generator matrix by the communication system. When the generator matrix is specified, the constraint length is also determined accordingly.

The monitor / notification unit 60 estimates the communication state of the communication path according to the packet decoding status in the decoder 70, and notifies the communication partner change unit 50 of the estimation result. As a result, the communication partner changing unit 50 instructs the communication partner changing unit 50 to select the coding rate or the generation matrix most suitable for the communication state of the current communication path.

The decoder 70 detects whether there is a packet loss or not according to the received information packet and redundant packet, and recovers it if there is a packet loss. The lost packet recovery process is the same as the process of the lost packet recovery unit 24 of the first embodiment described above.

As described above, in the present embodiment, the instructions for setting the coding rate are notified to each other between the changing unit 50 and the monitor / notification unit 60 which communicate with each other. In order to guarantee the communication quality, the instruction to set the coding rate must be surely transmitted to the communication partner. Therefore, it is preferable to transmit the instruction to set the coding rate using TCP that guarantees the communication data.

Next, the control of the coding rate in the data communication apparatus of this embodiment will be described. Here, the operations of the monitor / notification unit 60 and the changing unit 50 of the data communication device will be described with reference to respective flowcharts.

FIG. 7 is a flowchart showing the operation of the monitor / notification unit 60 of the data communication device. As shown in FIG. 7, first, the monitor / notification unit 60 acquires from the decoder 70 the number z of information packets that have not been recovered within a fixed time (step SA1).

Next, it is determined whether z is larger than 0 (step SA2). If z is greater than 0, the process of step SA4 is executed. In the opposite case, the process proceeds to step SA3. In step SA4, an instruction is sent to the communication partner changing unit 50 to reduce the coding rate. That is, if there is an information packet that has not been recovered within a certain period of time, it is determined that the state of the communication path is poor, the communication partner reduces the coding rate, and the recovery capability of the lost packet is improved.

In step SA3, within a fixed time,
It is determined whether z is held at 0. If z is held at 0 within a certain time, that is, there was no information packet that was not recovered in the decoder 70 within a certain time. In this case, it can be determined that the state of the communication path is good. Therefore, the process proceeds to step SA5, and an instruction to set the coding rate c to a large value is transmitted to the communication partner changing unit 50. As a result, the communication partner can greatly control the coding rate, and the efficiency of communication can be improved.

On the other hand, if it is determined in step SA3 that z has become a value other than 0 within a certain period of time, the process returns to step SA1 and the next determination is performed.

Next, the control of the coding rate in the changing unit 50 will be described. FIG. 8 is a flowchart showing the operation of the changing unit 50. As illustrated, the changing unit 50 first instructs the encoder 40 to set the coding rate to a predetermined initial value (step SB1). In response to this, the encoder 40 uses the initial coding rate c to generate a redundant packet and transmits it to the communication partner together with the information packet.

Next, it waits for an instruction from the monitor / notification unit 60 of the communication partner (step SB2). Do not change the coding rate until the communication partner instructs you to change it. Then, when an instruction to change the coding rate is received from the monitor / notification unit 60 of the communication partner, the encoder 40 is instructed to increase or decrease the coding rate accordingly (step SB3).

As described above, the monitor / notification unit 60 determines whether or not there is an information packet that cannot be recovered in the decoder 70 within a fixed time. As a result of the determination, when all the information packets can be recovered within a fixed time, it is determined that the communication state of the communication path is good, and the communication partner is instructed to set a large coding rate. In response to this, the coding rate is largely controlled in the communication partner, and the transmission efficiency is improved. On the contrary, if all the information packets cannot be recovered within a certain time, it is determined that the state of the communication channel is not good, and the communication partner is instructed to set the coding rate to a small value. According to this, the coding rate is controlled to be small in the communication partner, and the ability to recover lost packets can be improved. Instead, the transmission efficiency is reduced.
That is, according to the communication system of the present embodiment, based on the estimated communication state of the communication path, a trade-off is made between the data transmission efficiency and the lost packet recovery capability, and the optimum code is maintained while maintaining the lost packet recovery capability. The rate of conversion can be controlled.

FIG. 9 is a diagram for explaining the control of the coding rate in this embodiment, and shows the change of the packet transmitted with the change of the coding rate. As shown in the figure, first, one redundant packet is generated for every three information packets. That is, the packet coding rate is 3 /
It is set to 4 (0.75). Then, by changing the coding rate, two redundant packets are generated for every three information packets. That is, the coding rate is 3/5 (0.
60).

In the example shown in FIG. 9, the changed coding rate is
Although it is lower than before the change, the number of redundant packets generated is larger for the same number of information packets. Therefore, although the communication efficiency is reduced by the change of the coding rate, the ability of recovering the lost packet is increased on the receiving side.

As described above, the probability that packet loss will occur changes depending on the status of the transmission path. Therefore, it is necessary to optimally set the packet coding rate according to the status of the transmission path. It should be noted that the state of the transmission path can be estimated, for example, in the processing example shown in FIG. 7, based on the number z of packets that have not been recovered within a certain time in the decoding unit on the receiving side. In the example of FIG. 7, for example, the monitor / notification unit 7 on the receiving side
In 0, if it is determined that there is an information packet that cannot be recovered within a certain time and the communication state of the communication path is not good, an instruction to reduce the coding rate is transmitted to the changing unit 50 on the transmission side. To do. In response to this, the change unit 5 on the transmission side
0 sends an instruction to reduce the coding rate to the encoding unit 40,
The coding rate is set low. For this reason, on the receiving side, the ability to recover lost packets is improved, so that communication quality can be maintained.

Next, optimum control of the constraint length or generator matrix in the communication system of this embodiment will be described.
In the present embodiment, in the state of the communication path, by adaptively controlling the constraint length or the generation matrix when the redundant packet is generated on the transmission side, the transmission efficiency is kept high while maintaining the ability to recover the lost packet. it can.

The control of the constraint length or generator matrix can be realized by the data communication devices 100-3 and 100-4 having the configuration shown in FIG. However, here, the monitor / notification unit 60
And the processing of the changing unit 50 is different. Also, the encoder 40
In accordance with an instruction from the changing unit 50, changes the constraint length for generating a redundant packet, or changes the generation matrix used to generate and use a redundant packet.

The processes of the monitor / notification unit and the changing unit will be described below with reference to FIGS. 10 and 11. FIG. 10 is a flowchart showing the processing of the monitor / notification unit 60. As shown in the figure, first, the number w of information packets transmitted within a certain period of time and the number x of lost information packets
And the number z of information packets that have not been recovered are acquired from the decoder (step SC1).

Next, p'is calculated based on p '= x / w (step SC2).

Then, the calculated p'and the number z of packets not recovered within a fixed time are notified to the communication partner (step SC3).

FIG. 11 is a flow chart showing the processing of the communication partner changing unit 50. As illustrated, the changing unit 50 first instructs the encoder 40 of the generator matrix to be used first (step SD1). Here, for example, a plurality of generator matrices to which the respective identification numbers are assigned are determined in advance by the communication system, and the respective generator matrices and the identification numbers associated with the generator matrices are notified to both data communication devices in communication, and the respective data Such information is stored in, for example, the encoder 40 or the changing unit 50 of the communication device. Therefore, if the changing unit 50 indicates the identification number to the encoder 40, the encoder 40 can generate a redundant packet using a desired generator matrix.

Next, the changing unit 50 waits for an instruction from the communication partner (step SD2). The generator matrix currently in use is used as it is until the change instruction is received.

Next, when p'and the number z of packets which have not been recovered within a fixed time are received from the communication partner, the optimum generator matrix is obtained based on these information (step S
D3). Then, it instructs the encoder 40 on the obtained optimum generator matrix.

The above-mentioned monitor / notification section 60 and change section 5
By the processing of 0, the ratio p ′ of the number x of packets lost to the number w of transmitted information packets and the number z of unrecovered information packets in the fixed time are acquired by the receiving side and sent to the transmitting side. The changing unit 50 on the transmission side obtains an optimum generator matrix based on these pieces of information. Then, since the optimum generator matrix is instructed to the encoder 40, the redundant packet is generated in the encoder 40 based on the optimum generator matrix. Note that the changing unit 50 uses a generator matrix whose loss recovery capability improves on the receiving side when the values of p ′ and z increase, and conversely, when the values of p ′ and z decrease, the loss recovery capability increases. An optimal generator matrix is derived based on the principle of using a generator matrix that places more importance on transmission efficiency.

In addition to controlling the optimal generator matrix,
It is also possible to control the constraint length. As the constraint length becomes longer, the number of information packets related to the redundant packet to be generated increases, and the recovery capability of the lost packet at the receiving side improves, but the amount of calculation for generating the redundant packet increases, so the processing load increases. To increase. Therefore, in the changing unit 50, according to the values of p ′ and z sent from the receiving side,
By appropriately controlling the constraint length, the recovery capability of lost packets and the processing load can be optimally controlled.

FIG. 12 is a diagram showing control of the constraint length in the communication system of this embodiment. As shown in the figure, first, a redundant packet is generated by an encoding process with a constraint length of 4. That is, at this time, in order to generate the redundant packet, the four information packets immediately before that are used. Next, the constraint length is changed to 7. That is, the seven information packets immediately before that are used to generate the information packet. Increasing the constraint length improves the ability of the receiving side to recover lost packets, while increasing the processing load on the encoder. Therefore, in the communication system of this embodiment, the constraint length is optimally controlled according to the communication state of the communication path. The constraint length is determined by the generator matrix used for encoding. Therefore, by optimally controlling the generator matrix, the constraint length also changes accordingly.

As described above, in the communication system of the present embodiment, the state of the communication channel can be estimated on the receiving side according to the loss and recovery status of the information packet related to the decoding process, and according to the estimation result. Send a change instruction to the sender. In response to this, the transmission side appropriately changes the coding rate, the constraint length, or the generation matrix according to the change instruction to generate the redundant packet, so that the redundant packet is always based on the optimum coding process according to the situation of the communication path. Can be generated. As a result, the maximum transmission efficiency can be realized while keeping the communication quality constant regardless of the state of the communication path.

Third Embodiment FIG. 13 is a block diagram showing the third embodiment of the communication system according to the present invention. Note that, in FIG. 13, only the data communication device 100-5 forming the communication system is shown. The communication system of this embodiment may include a plurality of data communication devices 100-5. As shown in FIG.
The data communication device 100-5 according to the present embodiment includes an encoder 4
0, the decoder 70, and the application unit 80.

In the data communication device 100-5 of this embodiment, the encoder 40 adaptively changes its own encoding rate, constraint length or generator matrix according to the reproduction status of the information packet in the decoder 70. To do. Data communication device 100
When the -5 communicates with the other party via the communication network, it exchanges information on the same communication route as the other party. Therefore, the communication state of the communication path can be estimated according to the reception status of the information transmitted from the other party. Therefore, in the data communication device 100-5 of the present embodiment, the decoder 70
The state of reproduction of the information packets in, for example, the number of information packets lost within a certain period of time, the number of information packets that have not been recovered, etc. are estimated, and the communication state of the communication path is estimated. The coding rate and generator matrix of the device 40 are optimally controlled. With this, as in the second embodiment of the present invention described above, the state of the communication channel is estimated according to the reproduction state of the information packet received by the communication partner, and the optimum coding rate or generation matrix Compared with the control method of receiving the instruction, each data communication device can independently estimate the communication state of the communication path, and the optimization process can be realized more easily. Further, it does not require a TCP connection for transmitting the control information between the two parties, which is effective in saving resources.

14 and 15 are flow charts showing the control of changing the coding rate and the generator matrix in the data communication apparatus 100-5 of this embodiment. Hereinafter, the operation of the data communication device 100-5 according to the present embodiment will be described with reference to these flowcharts.

FIG. 14 illustrates the encoder 40 as shown.
First sets the coding rate c to an initial value (step SE
1). Next, the encoder 40 decodes the number w of information packets that should have been transmitted, the number x of lost information packets, the number y of redundant packets received, and the number z of information packets that have not been recovered within a fixed time. It is acquired from the container 70 (step SE2). The number of information packets that should have been sent w
Is estimated by the total value of the number of information packets actually received and the number x of lost information packets.

The encoder 40, based on the acquired information,
c ′ and p ′ are calculated as follows (step SE3).

[0082]

[Equation 2] c '= w / (w + y) (2)

[0083]

[Equation 3] p '= x / w (3)

That is, c'is the ratio of the number w of information packets that should have been transmitted from the communication partner within a fixed time and the total value of this number w of information packets and the number y of received redundant packets. This ratio is the actual code rate during the measurement period.
Further, p ′ is the ratio of the number z of information packets that have not been recovered to the number w of information packets that should have been transmitted,
That is, it means the loss rate of information packets.

Next, it is judged whether or not the number z of information packets which have not been recovered within a fixed time is larger than 0 (step SE4). If z is greater than 0, the process proceeds to step SE9, and the coding rate c is decreased according to p '. On the other hand, when z is 0 or less, the process proceeds to step SE5.

In step SE5, the calculated coding rate c'is compared with the currently used coding rate c. If c'is larger than c as a result of the comparison, step S
Proceeding to the processing of E8, using the calculated coding rate c ′,
Encoding processing, that is, generation of redundant packets is performed. On the other hand, if c'is less than or equal to c as a result of the comparison, the process proceeds to step SE6.

In step SE6, it is determined whether or not a fixed time has passed since the coding rate c was changed. As a result of the determination, when a certain time has passed since the coding rate was changed, the process proceeds to step SE7, and the coding rate c is increased. On the contrary, when the fixed time has not elapsed since the change of the coding rate, the process returns to the process of step SE2 and the above process is repeated.

By the above processing, the encoder 40 is
The coding rate c can be optimally controlled according to the reproduction status of the information packet in the decoder 70. For example, when there is an information packet that has not been recovered within a certain period of time, the coding rate c is appropriately set according to the calculated loss rate p ′ of the information packet. At this time, it can be estimated that the communication state of the communication path is not good as the loss rate p ′ is large. Therefore, the coding rate c is set low, and the recovery capability of the information packet lost on the receiving side is improved. Further, the calculated coding rate c ′ can be compared with the actually used coding rate c, and the optimum coding rate c can be set according to the comparison result.

Next, the control of the generator matrix in the data communication apparatus of this embodiment will be described with reference to the flowchart shown in FIG. As shown in FIG. 15, first,
In the encoder 40, the generator matrix used first is initialized (step SF1). The initial coding matrix is appropriately selected from, for example, a plurality of generator matrices predetermined by the communication system.

Next, the number w of information packets that should have been transmitted within a certain period of time, the number x of lost information packets, and the number z of unrecovered information packets are acquired from the decoder 70 (step SF2).

Then, based on the acquired information, p '= x
/ W is calculated (step SF3). Note that the calculated p ′ is, as described above, the number of lost information packets x with respect to the total number w of information packets transmitted within a fixed time.
Of the packet, that is, the loss rate of the information packet.

Then, an optimum generator matrix is obtained according to the calculated p'and the number z of information packets that have not been recovered. For example, the loss rate p ′ of the information packet becomes large,
Alternatively, if the number z of unrecovered information packets increases, it can be estimated that the communication state of the communication path is not good.
A generator matrix is selected that can improve the recoverability of lost information packets. In the opposite case, that is, when it is estimated that the condition of the communication path is good, a generator matrix that improves communication efficiency is selected.

As described above, according to the present embodiment, in the data communication device, the communication state of the communication channel is estimated according to the decoding status of the information packet in the decoder 70, and the encoder 40 according to the estimation result. The coding rate, constraint length, or generator matrix in is optimally controlled. As a result, it is not necessary to transmit the change instruction to the other party between the two data communication devices that communicate with each other, the configuration of the communication system can be simplified, and resources can be saved.

Fourth Embodiment FIG. 16 is a block diagram showing the fourth embodiment of the communication system according to the present invention. In the communication system of this embodiment, different types of information data are encoded at different encoding rates and transmitted. As shown in FIG. 16, a data communication device 100-6 that constitutes the communication system of this embodiment.
Is composed of encoders 40-1 and 40-2 and an application unit 80. The data communication device 100-6 according to the present embodiment is, for example, a device having a data communication function on a communication network called a gatekeeper.

The encoders 40-1 and 40-2 respectively encode the information data from the application section 80. Then, the encoded information packet is transmitted to the data communication device of the communication partner via the communication network. The application unit 80 is, for example, a functional block having a routing function, acquires different types of information data input by the respective input means, and outputs the information data to the encoder 40-1 or 40-2.

In the data communication device 100-6 of this embodiment, for example, the information data input to the encoder 40-1 is an information packet of video data generated according to an image (video) signal, and the encoder 40-2 The information data input to is an information packet of audio data generated according to the audio signal. Encoders 40-1 and 40-
2 performs encoding processing on these different types of information data at different encoding rates, generates redundant packets in addition to information packets, and transmits them to the communication partner.

According to the present embodiment, the information data of different types are subjected to the encoding processing at the encoding rates suitable for the characteristics of the information data, so that the respective information data have the desired transmission characteristics. For example, communication can be performed at the transmission rate and the recovery rate of lost packets.

Fifth Embodiment FIG. 17 is a block diagram showing the fifth embodiment of the communication system according to the present invention. As shown in the figure, in the communication system of this embodiment, the data communication device transmits information packets and redundant packets through different communication channels.

As shown in FIG. 17A, the normal gatekeeper uses one communication port (PortX) to sequentially transmit the information packet and the redundant packet generated based on the information packet. However, in the communication system of this embodiment,
As shown in FIG. 7B, the gatekeeper uses two communication ports PortX and Y, PortX is used for communication of information packets, and PortY is used for communication of redundant packets. Therefore, PortX
Only information packets are transmitted from PortY, and from PortY,
Only redundant packets are transmitted.

By transmitting the information packet and the redundant packet to the different communication ports as in the present embodiment, the information packet can be received even by the data communication device which is not compatible with the communication system for packet coding. . In this case, in a data communication device that is not compatible, one communication port is used, and P shown in FIG.
The information packet transmitted from ortX may be received. In this case, since the redundant packet cannot be received, the information packet lost during the transmission cannot be recovered. That is, according to the communication system of the present embodiment, it is possible to receive an information packet even with a data communication device that does not support the packet coding process. In this case, the information packet lost due to the communication state of the communication path cannot be recovered and the quality of communication deteriorates, but data communication by packet coding is generally available without being limited by the type of data communication device. Is.

Sixth Embodiment FIG. 18 is a diagram showing a sixth embodiment of the communication system according to the present invention. The data communication device of the communication system of this embodiment can handle variable-length packets.

As shown in FIG. 18, when encoding a variable-length packet, the length of each information packet is different, so the above-described encoding method cannot cope with this. Therefore, in the data communication device of the present embodiment, a redundant packet is generated according to the maximum length of the information packet to be encoded.

As shown in the figure, when a plurality of information packets having different payload lengths are to be coded, a redundant packet is generated in accordance with the information packet having the largest payload length among these information packets. That is, the length of the redundant packet becomes equal to the maximum length of the information packet. In this case, in the short information packet, the remaining part is filled with predetermined data, for example, 0. When each information packet is transmitted, only the valid part of each payload is transmitted.

On the receiving side, the lost information packet is recovered by using the redundant packet and other information packets. In this case, with respect to other received information packets, the same data as that at the time of the encoding process on the transmission side is embedded in a portion having a short length in accordance with the maximum length to recover the lost information packet.

According to the communication system of this embodiment, the redundant packet is generated in accordance with the maximum length of the information packet to be coded. Therefore, the packet coding is performed even in the packet communication using the variable length information packet. It can be applied and the reliability of variable length packet communication can be improved.

Seventh Embodiment FIG. 19 is a block diagram showing the seventh embodiment of the communication system according to the present invention. The communication system of the present embodiment is X
An information packet is transmitted by adding an X cast header to a normal information packet for performing the cast (Xcast).

In the case of FIG. 19A, an X cast header is added between the UDP header and the payload. Also,
In the case of FIG. 7B, it is added to the RTP header extension part. In the following description, an information packet with an X cast header added is referred to as an X cast packet.

In the case shown in FIG. 19A, since there is no method for distinguishing it from a normal information packet to which an X cast header is not added, both data communication devices communicating with each other perform communication using the X cast packet. It can be used when it is known in advance. Further, in the case shown in FIG. 19B, this restriction is not imposed. Since the RTP application ignores the unsupported RTP header extension, as shown in FIG. 19B, even if the X cast is added to the RTP header extension, the RTP application cannot identify the RTP header extension in the communication device.
Since the TP header extension is ignored, the payload can be processed correctly.

Next, the X cast header will be described with reference to FIG. The X cast header is shown in FIG.
As shown in, the format of Xcast4 defined by InternetDraft is used in principle. The destination list area is shown in FIG.
It is decided as shown in (b). In that, BIT
The MAP area is equivalent to the Xcast4 header, and the CA
The LLIDENTIFIER LIST consists of a list of identifiers of calls including the receiver terminal.

Data communication between multiple points by the above-mentioned X cast packet will be described below. The multipoint data communication can be applied to, for example, a video conference system in which a plurality of participants perform an interface.

FIG. 21 shows an example of the configuration of a normal multipoint communication system. As shown in the figure, the multipoint communication system includes a multipoint control device 120 and a gatekeeper 140.
-1,140-2, and a plurality of terminals 160-1,160
It is composed of -2, 160-3 and 160-4.

The multipoint control unit 120 controls the flow of information in the entire communication system. For example, in the case of a user of a communication system or a video conference system, the multipoint control device 120 synthesizes the video / audio signals transmitted from the terminals of the participants of the conference and outputs a single video / audio signal. . An X cast packet is formed based on the video / audio signal output here. The Xcast packet contains a list of call identifiers for all receiving terminals.

The gatekeepers 140-1 and 140-2 are preset with the following gatekeepers. The gatekeepers 140-1 and 140-2 scan the recipient list in the received X-cast packet, and if there is a terminal connected to it, remove the X-cast header from the packet and convert it as a normal unicast packet. , Send to that terminal. Otherwise, it rebuilds the recipient list and sends it to the next gatekeeper. If the next gatekeeper does not exist, the received packet is ignored.

For example, the gatekeeper 140-1 receives the information packets from the terminals 160-2 and 160-2 and sends them to the multipoint control unit 120. When the gatekeeper 140-1 receives an Xcast packet from the multipoint control unit 120 or the next gatekeeper 140-2, the gatekeeper 140-1 scans the receiving terminal list of the Xcast header of the packet, and the terminal 160- One or one
If 60-2 is included, the packet from which the X cast header has been removed is transmitted to the terminal 160-1 or 160-2 designated by the receiving terminal list. The gatekeeper 140-2 is also the gatekeeper 1
The received packet is processed almost in the same manner as 40-1.

In the multipoint communication system shown in FIG. 21, only one multipoint control unit 120 is included. Therefore, all packets are transmitted and received between the multipoint control device 120 and the gatekeeper 140-1 connected thereto, and the traffic is concentrated. As a result, the information transmission efficiency decreases, and the gatekeeper 140
The load of -1 increases excessively.

Therefore, in order to disperse the traffic in the entire communication system and improve the efficiency of information transmission,
A multipoint communication system that uses multiple multipoint controllers is advantageous.

FIG. 22 shows an example of the configuration of a multipoint communication system in which multipoint control devices are arranged in a distributed manner. As shown, this multipoint communication system includes multipoint control units (hereinafter referred to as MCUs) 120-1 and 120-.
2, gatekeepers 140-1, 140-2, and a plurality of terminals 160-1, 160-2, 160-3, 160-4
And 160-5.

The MCU 120-1 has a gatekeeper 140.
-1, and the MCU 120-2 is connected to the gatekeeper 1
40-2 is connected. In addition, the gatekeeper 140
-1, the terminals 160-1 and 160-2 are connected, and the gatekeeper 140-2 is connected to the terminals 160-3 and 160-4.
And 160-5 are connected.

In the multipoint communication system of this embodiment, the MCUs 120-1 and 120-2 function similarly to the multipoint control unit 120 in the multipoint communication system shown in FIG. In addition, gatekeepers 140-1 and 14
0-2 also functions similarly to the gatekeeper shown in FIG.

However, in the multipoint communication system of this example, the terminals 160-3 and 160 are connected to the gatekeeper 140-2.
-4 and 160-5, the MCU 120-2 is connected. Therefore, the gatekeeper 140-2 is the terminal 1
The packets received from 60-3, 160-4 and 160-5 are passed through the gatekeeper 140-1 to the MCU 120-
1 to the received packet directly to the MCU 120-2. In the MCU 120-2, the packet sent from the gatekeeper 140-2 is processed, an X cast packet is generated, a list of call identifiers of all receiving terminals is added to the X cast header, and the packet is sent via the gatekeeper 140-2. And send it to another gatekeeper.

In each gatekeeper, the recipient list of the Xcast packet sent from another gatekeeper is scanned, and when the terminal connected to itself is included, the Xcast header is removed and the unicast Generate a packet and send it to the terminal. Otherwise, it sends the received Xcast packet to the next gatekeeper or other MCU.

As described above, according to the multipoint communication system of this embodiment, an X cast packet is generated,
By embedding the call identifier list of all receiving terminals in it, when the terminal connected to itself is designated by the gatekeeper, the X-cast packet is converted into a unicast packet and transmitted to the designated terminal. Since the X cast packets are transmitted one after another by the plurality of gatekeepers, the information data can be transmitted to all designated terminals in the communication system. Further, by providing a plurality of multipoint control devices in the multipoint communication system, for example, as shown in FIG. 23, when one MCU fails, an Xcast packet can be generated by the remaining MCUs. Can guarantee the reliability of.

It goes without saying that the above-mentioned packet coding process can be applied to the X cast packet of this embodiment. For example, in each terminal, a redundant packet is generated by packet coding processing according to the information packet and is transmitted to the gatekeeper together with the information packet. Further, each MCU generates an X cast packet based on the information data received from each terminal via the gatekeeper, and also generates a redundant X cast packet.
Then, the X-cast packet and the redundant packet generated accordingly are transmitted via the gatekeeper. Therefore, the gatekeeper generates a unicast packet from the X cast packet and also generates a redundant unicast and transmits the unicast packet to the designated terminal. Therefore, the terminal receiving the packet recovers the lost information packet based on the information packet containing the valid information and the redundant packet. As a result, communication quality can be improved in the multipoint communication system.

Eighth Embodiment FIG. 24 is a block diagram showing the eighth embodiment of the communication system according to the present invention. The packet coding process in each embodiment of the present invention described above can be realized by software or hardware. Therefore, in the present embodiment, a packet format and the like for performing the above-mentioned packet coding processing by dedicated hardware are provided.

As one method of performing the packet coding process by hardware, a convolutional layer is newly defined in the protocol stack as shown in FIG. 24 (a). Implement the convolutional layer in hardware. In this case,
As shown in 4 (b), a unique number is defined as a protocol in the Ethernet header. The payload is redundant data generated by the IP header and the datagram or the convolutional code. The convolutional header also includes a sequence number indicating the order of packets, identification information indicating information / redundant packet discrimination, upper layer information, and the like.

According to this embodiment, the packet coding process can be realized by hardware instead of software, so that the coding process can be speeded up. By applying the hardware-based packet coding apparatus to a communication system, particularly a communication system that requires high-speed data transmission such as a video conference, communication quality, response characteristics, etc. can be significantly improved.

Ninth Embodiment FIG. 25 is a block diagram showing the ninth embodiment of the communication system according to the present invention. The communication system of the present embodiment extends the packet coding process to the Galois field. In this case, the packet is regarded as a concatenation of a plurality of symbols on the Galois field. That is, each q bit in the packet is GF (2 ^q )
Considered to be the origin of the above. A redundant q bit is generated for each q bit, and a redundant packet is generated based on this. Further, the coefficient of the generator matrix is also an element on GF (2 ^q ). Here, q is an integer larger than 1 and smaller than the bit length of the packet.

As shown in FIG. 25, packet data is constructed based on the information packet together with the redundant packet formed on the Galois field.

[0129]

As described above, according to the communication device, the communication system and the video conference system using the same of the present invention, the packet coding process is performed on the information packet by a predetermined coding method to obtain the information. A redundant packet is generated based on the packet, and both the information packet and the redundant packet are transmitted on the communication network. When an information packet loss occurs in the middle of the transmission path, the lost information packet can be recovered based on the redundant packet and other information packet received on the receiving side. As a result, communication quality can be improved by adding a simple configuration to the communication device. Further, according to the present invention, in a multipoint communication system such as a video conference system, by disposing the multipoint control devices in a distributed manner, concentration of traffic can be avoided, and the reliability of the communication system can be guaranteed.
Further, according to the present invention, there is an advantage that a high-quality and high-speed communication system and a video conference system can be realized by implementing the packet coding processing by hardware.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a communication system according to the present invention.

FIG. 2 is a conceptual diagram showing generation of redundant packets in packet encoding.

FIG. 3 is a diagram showing a redundant packet generation process based on an information packet.

FIG. 4 is a diagram showing an example of an information packet and a redundant packet generated based on the information packet.

FIG. 5 is a schematic diagram for explaining a lost packet recovery process.

FIG. 6 is a configuration diagram showing a second embodiment of a communication system according to the present invention.

FIG. 7 is a flowchart showing processing of a monitor / notification unit of the data communication device.

FIG. 8 is a flowchart showing processing of a changing unit.

[Fig. 9] Fig. 9 is a diagram showing a change in a coding rate in the present embodiment.

FIG. 10 is a flowchart showing processing of a monitor / notification unit.

FIG. 11 is a flowchart showing processing of a changing unit.

FIG. 12 is a diagram showing control of a constraint length in the communication system of the present embodiment.

FIG. 13 is a configuration diagram showing a third embodiment of a communication system according to the present invention.

FIG. 14 is a flowchart showing control of changing the coding rate in the data communication apparatus according to the present embodiment.

FIG. 15 is a flowchart showing change control of a generation matrix in the data communication device of the present embodiment.

FIG. 16 is a configuration diagram showing a fourth embodiment of a communication system according to the present invention.

FIG. 17 is a configuration diagram showing a fifth embodiment of a communication system according to the present invention.

FIG. 18 is a configuration diagram showing a sixth embodiment of a communication system according to the present invention.

FIG. 19 is a configuration diagram showing a seventh embodiment of a communication system according to the present invention.

FIG. 20 is a diagram showing an X cast header.

FIG. 21 is a configuration diagram showing an example of a multipoint communication system.

FIG. 22 is a configuration diagram showing an example of a multipoint communication system in which multipoint control devices are arranged in a distributed manner.

FIG. 23 is a diagram showing a failure situation in a multipoint communication system in which multipoint control devices are arranged in a distributed manner.

FIG. 24 is a configuration diagram showing an eighth embodiment of the communication system according to the present invention.

FIG. 25 is a configuration diagram showing a ninth embodiment of a communication system according to the present invention.

[Explanation of Codes] 10 ... Packet Encoding Device, 20 ... Packet Reproducing Device,
120, 120-1, 120-2 ... Multipoint control device (M
CU), 140-1, 140-2 ... Gatekeeper, 16
0-1, 160-2, 160-3, 160-4, 160
-5 ... a terminal.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Mamoru Ohara             1-1 Minami Osawa, Hachioji City, Tokyo             Tokyo Metropolitan University (72) Inventor Ryu Suzuki             1-1 Minami Osawa, Hachioji City, Tokyo             Tokyo Metropolitan University (72) Inventor Satoshi Fukumoto             1-1 Minami Osawa, Hachioji City, Tokyo             Tokyo Metropolitan University (72) Inventor Kazuhiko Iwasaki             1-1 Minami Osawa, Hachioji City, Tokyo             Tokyo Metropolitan University F-term (reference) 5C064 AA02 AB04 AD08 AD09 AD14                       AD16                 5K014 BA10 CA02 FA11                 5K015 AB01 JA04 JA05

Claims (22)

[Claims] 1. Redundant packet generating means for generating a redundant packet according to a predetermined encoding method according to an information packet to be transmitted, and a recovery corresponding to the encoding method when a packet loss is detected in a received packet. And a packet recovery means for recovering the lost packet based on the redundant packet and the information packet received by the method. 2. The communication device according to claim 1, wherein the redundant packet generation means controls the coding rate according to the occurrence status of the packet loss to generate the redundant packet. 3. The communication apparatus according to claim 1, wherein the redundant packet generating means changes the generating matrix used for the encoding process to generate the redundant packet in accordance with the occurrence status of the packet loss. 4. The communication device according to claim 1, wherein the redundant packet generating means changes the constraint length of the encoding process to generate the redundant packet in accordance with the occurrence status of the packet loss. 5. The communication apparatus according to claim 1, wherein the redundant packet generation means sets different coding rates for different types of information data to be transmitted and performs coding processing. 6. The communication device according to claim 1, wherein said redundant packet generating means generates a redundant packet for the variable length packet in accordance with the maximum packet length of the information packets to be encoded. 7. The communication device according to claim 6, wherein said redundant packet generation means embeds predetermined data in the remaining bits of the information packet shorter than the maximum packet length to make the maximum packet length. 8. The communication device according to claim 1, further comprising transmitting means for transmitting the information packet and the redundant packet through different communication channels. 9. The communication device according to claim 1, further comprising control means for adding a destination list including a plurality of destinations immediately after the UDP header in the UDP packet including the information packet and the redundant packet. 10. The communication device according to claim 1, further comprising control means for adding a destination list including a plurality of destinations to an extended portion of the RTP header in the UDP packet including the information packet and the redundant packet. 11. The communication device according to claim 1, wherein the encoding process in said information packet generating means is executed by hardware. 12. The communication apparatus according to claim 1, wherein said redundant packet generating means encodes a packet by regarding it as a concatenation of a plurality of symbols on a Galois field. 13. Redundant packet generating means for generating a redundant packet by a predetermined encoding method according to an information packet to be transmitted, and recovery when the packet loss is detected in a received packet, the restoration corresponding to the encoding method. A communication system for transmitting the information packet and the redundant packet via a communication network using a communication device having a packet recovery means for recovering the lost packet based on the redundant packet and the information packet by a method. 14. A communication control means, which is connected to a plurality of the communication devices, adds a destination list indicating a plurality of destinations to a packet transmitted by each communication device, and controls the destinations of the packets. The communication system according to claim 13. 15. The communication system according to claim 14, further comprising relay means for transmitting the packet to a plurality of communication devices based on the destination list included in the packet. 16. The communication system according to claim 14, wherein at least two communication control means are provided, and when one of the communication control means fails, another communication control means controls the communication. 17. An information packet is generated based on predetermined information data and is transmitted via a communication network, and
A terminal that reproduces information data according to an information packet transmitted via the communication network and a terminal connected to a plurality of the terminals via a communication network, and communicates the information packet transmitted from the plurality of terminals. The relay means for transmitting to the control means, and the communication means add a destination list indicating a plurality of destinations to a predetermined area of the packet transmitted from the relay means, and transmit to the relay means. Conference system. 18. The relay means, when a terminal connected to itself is designated by the destination list in the received packet, transmits the packet to the designated terminal. Video conferencing system. 19. The relay means transmits the packet to another relay means when a terminal connected to itself is not designated by the destination list in the received packet. Video conferencing system. 20. The video conference system according to claim 17, wherein at least two communication control means are provided, and when any one of the communication control means fails, the other communication control means controls the communication. 21. The terminal generates a redundant packet according to a predetermined encoding method based on the information packet,
18. The video conference system according to claim 17, wherein the video conference is transmitted together with the information packet to the relay means. 22. The communication control means generates a redundant packet according to a predetermined encoding method based on the information packet sent from the relay means, and transmits the redundant packet to the relay means together with the information packet. Item 17. The video conference system according to Item 17.