JP2003179640A - Supplementing system and method for missing packet in broadcast communication - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To supplement a missing packet in UDP/IP broadcast communication without putting loads on a data distribution server. <P>SOLUTION: By having a distribution server 2 sequentially transmit distribution data by UDP/IP broadcast by packets to clients 3 sharing the packets of the broadcast divided into respective groups and sequentially received for the respective groups and saving them separately by sequence number units, the client 3 requests the missing packet to the other client in the case of detecting the missing packet while receiving the packets and the other client transmits the missing packet to the client 3 of a request origin instead of a server when the requested missing packet is saved. <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 missing packet complementing method and method in broadcast communication, and more particularly to a missing packet complementing method and method in UDP / IP broadcast communication from a server that distributes data.

[0002]

2. Description of the Related Art Conventionally, UD by UDP / IP packet
When performing data distribution by P / IP broadcast communication, UDP / IP packets may not be received depending on the load condition of the network or the load condition of each data receiving client. Furthermore, UDP / IP communication is
Even if a packet loss occurs, it does not have a mechanism for retransmitting, so it is not possible to reacquire a UDP / IP packet that could not be received. Further, the UDP / IP broadcast communication does not have a mechanism for complementing a packet lost by retransmission even if a packet loss occurs.

Therefore, a technique for complementing a lost packet in broadcast communication has come out. For example, referring to FIG. 14, a conventional technique for complementing a lost packet in the broadcast communication is a LAN 4 which is a local area network, a data delivery server 5 which delivers data by UDP / IP broadcast by a UDP / IP packet, It is composed of a plurality of data receiving clients 6. From the data distribution server 5,
By giving a sequence number to the packet to be distributed to each data receiving client 6, the data receiving client 6 on the receiving side detects a packet loss and requests the data distribution server 5 to transmit the packet. . A conventional technique for complementing a lost packet in broadcast communication is disclosed in, for example, Japanese Patent Laid-Open No. 6-252.
Also disclosed in Japanese Patent No. 897 is a technique in which "the transmitting side broadcasts to the receiving side by UDP / IP with a sequence number and requests the retransmission to the transmitting side when the receiving side detects a missing number". Has been done.

[0004]

However, in the conventional technique for complementing a lost packet in broadcast communication, the data receiving client (reception side) can detect the loss of the packet, but a request for resending the lost packet is issued. If the data distribution server (transmission side) is used, there is a problem in that the data distribution server itself is also overloaded. This causes a problem that the data delivery speed to other data receiving clients decreases.

In view of the above problems, an object of the present invention is to provide complementation of a lost packet in UDP / IP broadcast communication without imposing a load on the data distribution server.

[0006]

In order to achieve the above object, in the first method of complementing a lost packet in broadcast communication of the present invention, a server packets delivery data to a plurality of clients via a network. In the method of complementing a lost packet in broadcast communication for transmitting in sequence by UDP / IP broadcast by the method, if the client detects a lost packet while receiving the packet from the server, the lost packet is sent to another client. Requesting means for making the request, and sending means for sending the missing packet to the request source instead of the server if the other client has saved the requested missing packet. .

Further, in the above-mentioned method of complementing a lost packet in the first broadcast communication, the packets received in sequence by the clients divided into groups are shared for each group and are divided and stored in sequence number units. The requesting means has a storage means, and the requesting means makes a request for the lost packet by UDP / IP broadcast to the other client by the client detecting the lost packet.

Further, in the method of complementing a lost packet in broadcast communication, storage means for sharing the packets received by the clients divided into groups in sequence for each group, and storing the packets in sequence number units. Management means for managing destination information for each group based on the storage means in a table, and the request means selects a client to be requested from the table by the client detecting the lost packet. It is characterized in that the lost packet is requested to the selected client.

A second method of complementing a lost packet in broadcast communication according to the present invention is a broadcast communication in which a server transmits distribution data by packets in a UDP / IP broadcast sequence to a plurality of clients via a network. In the method of complementing a lost packet, the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a provision means for imparting a group number to the clients divided into respective groups is provided, and the client is Determination means for determining the type,
If the determining means determines that the data is the distribution data from the server, whether the sequence number is complete by comparing the sequence number of the packet received this time with the sequence number of the packet received last time from the server. A detection means for checking and detecting a missing packet,
When the detecting means detects a missing sequence number, the requesting means for resending the lost packet to all the clients connected to the network, and the detecting means must detect the missing sequence number. When the sequence number of the received packet belongs to its own group number, a saving unit that saves the received packet, and when the determining unit determines that the request to retransmit the lost packet, the requested sequence number is the own sequence number. And a processing unit which, on behalf of the server, retransmits the requested lost packet when belonging to the group number.

The third method of complementing a lost packet in broadcast communication according to the present invention is a method of complementing a missing packet in broadcast communication, in which a server sends UDP / IP broadcast distribution data as packets to a plurality of clients via a network. In the complementing method of the lost packet in the broadcast communication to be transmitted in a sequence in, the plurality of clients are divided into groups corresponding to the sequence number of the packet, and a provision means for imparting a group number to the clients divided into respective groups, The client is a determination means for determining the type of the received packet,
If the determining means determines that the data is the distribution data from the server, whether the sequence number is complete by comparing the sequence number of the packet received this time with the sequence number of the packet received last time from the server. A detection means for checking and detecting a missing packet,
When a missing sequence number is detected by the detecting means, an extracting means that extracts a group number corresponding to the missing sequence number, and a first client that selects one client that belongs to the extracted group number The selection means, the first retransmission request means for making a request to retransmit the lost packet to the client selected by the first selection means, and the detection means when the missing of the sequence number is not received, When the sequence number of the packet belongs to its own group number, a saving unit that saves the received packet, and when the determination unit determines that the lost packet is a retransmission request, saves the requested lost packet. And a resending unit that resends the requested lost packet on behalf of the server when requested. That.

Further, in the above-mentioned missing packet complementing method in the third broadcast communication, when the judgment unit judges that the missing packet is to be retransmitted, the requested missing packet is not saved. Second selecting means for selecting one from the clients belonging to the same group as the group number, and second retransmission requesting means for requesting retransmission of the lost packet to the client selected by the second selecting means. It is characterized by having.

Further, in the above-mentioned method of complementing a lost packet in the second or third broadcast communication,
When the determination unit determines that the lost packet is to be retransmitted, if the lost packet has not been received, it includes a capture unit that captures the lost packet.

The first method of complementing a lost packet in broadcast communication according to the present invention is a method for broadcast communication in which a server transmits a distribution data in a packet by UDP / IP broadcast in sequence to a plurality of clients via a network. A method of complementing a missing packet, wherein when the client detects a missing packet while receiving the packet from the server, the other client requests the missing packet, and the other client requests the missing packet. If the stored lost packet is stored, the lost packet is transmitted to the request source instead of the server.

Further, in the above-mentioned method of complementing a lost packet in the first broadcast communication, the packets received by the clients divided into groups in a sequence are shared for each group and stored in sequence number units. When the client that has detected the lost packet requests the lost packet, the other packet is requested by the UDP / IP broadcast for the lost packet.

Further, in the above-mentioned method of complementing a lost packet in the first broadcast communication, the packets received in sequence by the clients divided into groups are shared by the groups and are divided and stored in sequence number units. , The client manages the destination information for each group based on the storage means in a table, and when the client that detects the lost packet requests the lost packet, selects the client to be requested from the table, It is characterized in that the lost packet is requested to the selected client.

The second method of complementing a lost packet in broadcast communication according to the present invention is a method of broadcast communication in which a server transmits distribution data in packets by UDP / IP broadcast in sequence to a plurality of clients via a network. A method of complementing a lost packet, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is given to the clients divided into respective groups, and the client is
When the packet is received from the server, it is checked whether there is a missing sequence number, and if a missing sequence number is detected, a request to retransmit the missing packet is sent to all the clients connected to the network. If the sequence of the sequence number is normal and the sequence number of the received packet belongs to the own group number, the received packet is saved, and when the missing packet is received, the requested sequence number is the own group number. The missing packet requested to belong to the number is transmitted to the request source instead of the server.

A third method of complementing a lost packet in broadcast communication according to the present invention is a broadcast communication in which a server transmits distribution data by packets in a UDP / IP broadcast sequence to a plurality of clients via a network. A method of complementing a lost packet, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is given to the clients divided into respective groups, and the client is
When the packet is received from the server, the sequence number is checked for omissions, and if omission of the sequence number is detected, the group number corresponding to the omission sequence number is extracted, and the extracted group number is added to the extracted group number. One of the clients that belong to the selected client is selected, a request for resending a lost packet is made to the selected client, and if the sequence number is normal, the received sequence number of the received packet belongs to the own group number. When a packet is saved and a request to retransmit the lost packet is received, if the requested lost packet is saved, the requested lost packet is transmitted to the request source on behalf of the server. It has a feature.

Further, in the above-mentioned second or third method of complementing a lost packet in broadcast communication, when the client receives the lost packet, if it has not already received, the client fetches the lost packet and already receives it. If there is, the lost packet is abandoned.

A fourth method of complementing a lost packet in broadcast communication according to the present invention is a broadcast communication in which a server transmits distribution data in a packet by UDP / IP broadcast to a plurality of clients via a network in sequence. A method of complementing a lost packet, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is given to the clients divided into respective groups, and the client is
When the type of the received packet is determined and the type of the received packet is determined to be distribution data from the server, the sequence number of the packet received this time and the sequence number of the packet previously received from the server By comparing the above, the sequence number is checked for missing packets and a missing packet is detected.If the missing sequence number is detected by checking for missing sequence numbers, a request to retransmit the missing packet is connected to the network. If the sequence number is normal when the sequence number is normal, the received packet is saved if the sequence number of the received packet belongs to its own group number. , The sequence number of the received packet belongs to its own group After checking whether or not the packet sequence number received this time is replaced with the sequence number of the previously received packet, and the type of the received packet is determined to be a retransmission request for the lost packet, the requested sequence number is When the packet belongs to the group number, the requested lost packet is retransmitted on behalf of the server.

Further, a fifth method of complementing a lost packet in broadcast communication according to the present invention is a broadcast communication in which a server transmits a distribution data in a packet by UDP / IP broadcast in sequence to a plurality of clients via a network. A method of complementing a lost packet, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is given to the clients divided into respective groups, and the client is
When the type of the received packet is determined and the type of the received packet is determined to be distribution data from the server, the sequence number of the packet received this time and the sequence number of the packet previously received from the server By comparing the sequence numbers, the sequence numbers are checked for missing packets to detect missing packets.If the missing sequence numbers are detected by checking for missing sequence numbers, the group number corresponding to the missing sequence number is detected. The client selects one client that belongs to the extracted group number, requests the resending of the lost packet to the selected client, and checks the sequence number to see if the sequence number is correct. And the sequence number of the received packet belongs to the own group number. The save the received packet to,
After checking whether or not the sequence number of the received packet belongs to its own group, the packet sequence number received this time is replaced with the sequence number of the previously received packet, and the type of the received packet is determined to be the retransmission request for the lost packet. In this case, if the requested lost packet is stored, the requested lost packet is retransmitted on behalf of the server.

Further, in the fifth method of complementing a lost packet in broadcast communication, when it is determined that the type of the received packet is a retransmission request for the lost packet, the requested lost packet is saved. If not, one client is selected from the clients belonging to the same group as the own group number, and the request to retransmit the lost packet is made to the client selected by the second selecting means.

Further, in the above-mentioned fifth method of complementing a lost packet in broadcast communication, when it is determined that the type of the received packet is retransmission of the lost packet, the lost packet is already received if the lost packet has not been received. Is taken in and the lost packet is abandoned if the lost packet has already been received.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION Next, referring to FIG. 1 for explaining a first embodiment of the present invention with reference to the drawings,
In the first embodiment of the present invention, a communication line such as LAN1 which is a local area network, a data distribution server 2 connected to LAN1, and a plurality of data receiving clients 3 (3-1-1) connected to LAN1. ~ 3-1
5 ... 3-10-1 to 3-10-5). The UDP / IP protocol is used as the communication protocol between the data distribution server 2 and the plurality of data receiving clients 3.

The data receiving clients 3 are each configured as a group (in the example, groups 0 to 9). That is, for convenience of explanation of the present invention, for example,
The numbers in the middle of the data receiving clients 3-1-1 to 3-1-5 indicate the group 1, the data receiving clients 3-2-1 to 3-2-5 indicate the group 2, and the data receiving client 3- 10-1 to 3-10-5 indicate the group 0. Further, the last number given to the data receiving client 3 indicates the identification number of the data receiving client 3 in each group for the convenience of explanation of the present invention.

Referring to FIG. 2, the data distribution server 2
Is an information processing device such as a server, a workstation, or a personal computer, and is a processor (CPU) not shown.
Control unit 21 which operates under program control by means of a LAN
An interface unit 22 for exchanging data with other devices via 1 and a storage unit 23 for storing information such as distribution data to be transmitted to the data receiving client 3.

The storage unit 23 in FIG. 2 is a readable / writable recording medium device (for example, a magnetic disk).

Referring to FIG. 4, there is shown a functional block of a processing operation of a program executed by the control unit 21 of the data distribution server 2 of FIG. 2 when transmitting distribution data to the data receiving client 3, and the control unit thereof. Reference numeral 21 denotes a data creation unit 211 that creates distribution data in the data reception client 3, and a distribution processing unit 21 that stores the packet type in a packet carrying the created distribution data, assigns a sequence number, and increments the sequence number.
2 and a transmission unit 213 that performs data distribution by UDP / IP broadcast communication via the interface unit 22.

Referring to FIG. 3, the data receiving client 3 is an information processing device such as a workstation or a personal computer, and includes a control unit 31 operated by a processor (CPU) under program control and another device via the LAN 1. An interface unit 32 for exchanging data with
The management table 33 that manages the own group number, the previous sequence number of the received packet, etc., accompanying UDP / IP packet communication, and the last digit of the sequence number of the received packet from the data distribution server 2 matches the own group number. In this case, the packet storage unit 34 stores the received packet.

The management table 33 and the packet storage unit 34 of FIG. 3 are readable and writable memories (for example, R
It is assigned to an area within AM).

Referring to FIG. 6, the management table 3 of FIG.
3 shows an example of the configuration in 3, the own IP address indicating the IP address of the own device, the own group number indicating the group number of the own device, and the sequence number belonging to the own group number (multiple lower digits can be set). And a previous sequence number indicating the latest sequence number (used as the previously received sequence number) of the UDP / IP packet received for each transmission source (for example, IP address) A request for retransmission of a lost packet is being made, and a request for retransmission is being made. In this case, the self IP address, the self group number, and the storage sequence number are stored in the data receiving client 3 from an input device (for example, a keyboard) by a maintenance person or the like in advance in a nonvolatile recording medium device (not shown). It is set in an area (for example, a magnetic disk), and when the data receiving client 3 starts up, the recording medium device sets its own IP address in the area of its own IP address in the management table 33 to the area of its own group number. The self group number has the lower digits of a plurality of sequence numbers in the storage sequence number area (in this description, only one lower digit is set)
Are copied respectively. It should be noted that it is stored in the data distribution server 2 in advance, and when the data receiving client 3 starts up, the data distribution server server 2 automatically sends its own IP address, group number, and storage sequence number. Alternatively, it may be written in the management table 33.

Referring to FIG. 5, there is shown a functional block of a processing operation of a program executed by the control unit 31 of the data reception client 3 of FIG. 3 when receiving a UDP / IP packet. A reception unit 311 that receives a packet via the interface unit 32, a reception packet type determination unit 312 that determines the type of the received packet, and a sequence number of the packet that is received this time compared with the sequence number of the packet that was received last time. Missing packet detector 313 that checks for missing packets and detects missing packets, and missing packet resend request processor 314 that makes a missing packet resend request when missing packets are detected.
And a data saving processing unit 315 that saves data when the last digit of the sequence number of the distribution packet from the data distribution server 2 belongs to its own group number, and a corresponding process that performs processing according to the data content of the received packet. Part 316
A missing packet receiving unit 317 that performs processing when the type of the received packet is a missing packet resend; and a data delivery server 2 that resends the requested missing packet if the type of the received packet is a missing packet resend request. And a lost packet retransmission processing unit 318 which is performed instead of the above. Figure 5
The data storage processing unit 315 shown in (1) performs the following operation depending on the group of the data receiving clients 3.
Group 1 data receiving clients 3-1-1, 3-
In the data storage processing units 315 of 1-2, 3-1-3, 3-1-4, 3-1-5, the last digit of the sequence number of the received UDP / IP packet is 1 (the storage sequence of the management table). 1 is set in the number area), the packet is stored in the packet storage unit 34. Group 2 data receiving clients 3-2-1, 3-2-2, 3-2
-3, 3-2-4, 3-2-5 data storage processing unit 31
If the last digit of the sequence number of the received UDP / IP packet is 2 (2 is set in the storage sequence number area of the management table), 5 stores the packet in the packet storage unit 34. Group 3 data receiving clients 3-3-1, 3-3-2, 3-3-3, 3-3-
The data storage processing units 315 of 4, 3-3-5 receive UDs.
If the last digit of the sequence number of the P / IP packet is 3 (3 is set in the storage sequence number area of the management table), the packet is stored in the packet storage unit 34. Group 4 data receiving client 3-4-
1,3-4-2,3-4-3,3-4-4,3-4-5
If the last digit of the sequence number of the received UDP / IP packet is 4 (4 is set in the storage sequence number area of the management table), the data storage processing unit 315 of
The packet is stored in the packet storage unit 34. Group 5 data receiving clients 3-5-1, 3-5
In the data storage processing units 315 of 2, 3-5-3, 3-5, 3-5-5, the last digit of the sequence number of the received UDP / IP packet is 5 (the storage sequence number area of the management table). Is set to 5), the packet is stored in the packet storage unit 34. Group 6 data receiving clients 3-6-1, 3-6-2, 3-6-
3, 3-6-4, 3-6-5 data storage processing unit 315
If the lower one digit of the sequence number of the received UDP / IP packet is 6 (6 is set in the storage sequence number area of the management table), the packet is stored in the packet storage unit 34. Group 7 data receiving clients 3-7-1, 3-7-2, 3-7-3, 3-7-
The data storage processing units 315 of 4, 3-7-5 receive UDs.
If the last digit of the sequence number of the P / IP packet is 7 (7 is set in the storage sequence number area of the management table), the packet is stored in the packet storage unit 34. Group 8 data receiving client 3-8-
1, 3-8-2, 3-8-3, 3-8-4, 3-8-5
If the last digit of the sequence number of the received UDP / IP packet is 8 (8 is set in the storage sequence number area of the management table), the data storage processing unit 315 of
The packet is stored in the packet storage unit 34. Group 9 data receiving clients 3-9-1, 3-9-
In the data storage processing units 315 of 2, 3-9-3, 3-9-4, and 3-9-5, the last digit of the sequence number of the received UDP / IP packet is 9 (the sequence number area for storing the management table). 9 is set), the packet is stored in the packet storage unit 34. Group 0 data receiving clients 3-10-1, 3-10-2, 3-1
In the data storage processing units 315 of 0-3, 3-10-4, 3-10-5, the last digit of the sequence number of the received UDP / IP packet is 0 (0 is set in the storage sequence number area of the management table). Stored), the packet is stored in the packet storage unit 34.

Referring to FIG. 7, the format (or UDP) of the UDP segment in the packet used for exchanging data accompanying UDP / IP broadcast communication.
Datagram), the format of which is the Internet Architecture Committee (IA).
B: Internet ArchitectureBo
ard) is RFC (Request for Comm)
UDP header area and UDP specified by ents)
The UDP data area includes a packet type storage area, a sequence number area, and a data area according to the present invention. In this packet type storage area, whether the relevant packet is a delivery packet from the data delivery server, a missing packet resend request requesting resending of a missing packet, or a missing packet resend sent by a missing packet resend request An identification code for distinguishing between is stored. Next, with reference to FIGS. 1 to 8, the overall operation of the present embodiment will be described centering on the flowchart of FIG. Note that FIG. 8 shows the UDP /
7 is a flowchart showing the operation of the control unit 31 when receiving an IP packet.

Now, the data distribution server 2 tries to transmit the data in the storage unit 23 to the all data receiving clients 3 in order by dividing the data several times. Then, in each data reception client 3, the control unit 31 controls the data creation unit of the control unit 21 of the data distribution server 2 of the management table 33 to 21
1 reads the data from the storage unit 23 and uses the UDP / I
Since it is sent by P broadcast communication, the delivery processing unit 21
Pass control to 2. The delivery processing unit 212 sets necessary information (port number of destination and source, etc.) in the UDP header area of the UDP segment of FIG. 7, and also assigns a packet type (delivery packet in this case) and a sequence number. Further, the delivery processing unit 212 sets the destination of the packet to the multicast address, sets the transmission source to the own IP address read from the management table 33, and passes control to the transmission unit 213. In this case, the delivery processing unit 212 controls the transmission unit 213 to send the divided data by placing it in the data storage area of the UDP segment.
Each time it is passed to the, the sequence number in the packet is incremented sequentially from 0. The transmission unit 213, to which the control is passed,
Through the interface unit 22, packets are continuously sent to all the data receiving clients 3 in order.
That is, the data distribution server 2 transmits the distribution packets to the all data reception clients 3 in sequence from the sequence number 0 by the UDP / IP broadcast communication. The control unit 31 of each data reception client 3 which has received the distribution packet at the reception unit 311 via the interface unit 32 determines the type of the received packet at the reception packet type determination unit 312 (step A in FIG. 8).
0, A1, A2), a packet distributed from the data distribution server 2 (in this case, the content of the packet type storage area in FIG. 7 is a distribution packet), the missing packet detection unit 3
At 13, it is checked whether or not there is any omission in the sequence number received this time from the data distribution server 2. The sequence number of the received packet (in this case, the contents of the sequence number storage area in FIG. 7 is meant. (Omitted) and the previous sequence number (corresponding to the data distribution server 2 that is the transmission source) stored in the management table 33 are compared (step A3 in FIG. 8). Note that step A3
In the above, the lost packet detection unit 313 of the control unit 31
If it is the first sequence number (for example, 001), it is determined that the sequence number is normal because the transmission source management server 2 is not registered in the management table 33. Missing sequence number (This sequence number is +2 more than last time)
The control unit 31 of the data receiving client 3 which has detected the above value) transmits the lost packet retransmission request using UDP / IP broadcast communication to all the data receiving clients 3 in the lost packet retransmission processing unit 314. Then (step A7 in FIG. 8), data complement is attempted. In this case, the lost packet retransmission processing unit 314 sets the destination in the packet to the multicast address and reads the transmission source from the management table 33.
The address is set, and the type code of the missing packet retransmission request is added to the packet type storage area of FIG. 7, the sequence number requested to the sequence number storage area is added to the IP header of the lost packet, and all the The data is transmitted to the data receiving client 3 via the interface unit 32, and the request for retransmission of a lost packet in the management table 33 is turned “ON”. After the processing of step A3 or step A7, in the data storage processing unit 315 of the control unit 31, the value of the last digit of the sequence number of the received packet belongs to the own group number stored in the management table 33 (see FIG. 8). When the received packet is stored in the packet storage unit 34 (step A4 in FIG. 8).
5), the previous sequence number stored in the management table 33 is replaced with the received sequence number (step A6 in FIG. 8), and the corresponding processing unit 316 performs processing according to the data content of the received packet (FIG. 8). Step A
7). In step A6, if the sequence number is the first (for example, 001), the data storage processing unit 315 does not register the IP address of the data distribution server 2 of the transmission source in the management table 33, and thus the transmission source area The IP address of the data distribution server 2 is set in the previous sequence number area, and the sequence number (eg, 001) of the received packet is set in the previous sequence number area. Also, step A
In 6, when the sequence number is the last, the data storage processing unit 315 deletes the information (transmission source, previous sequence number, retransmission request in progress) corresponding to the transmission server 2 of the transmission source from the management table 33.

In step A4, if the last digit of the sequence number of the received packet does not belong to its own group number, the control unit 31 does not store the received packet and the data storage processing unit 315 does not store the received packet. , Management table 33
Replace the previously stored sequence number with the received sequence number (step A6 in FIG. 8), and the corresponding processing unit 3
At 16, processing is performed according to the data content of the received packet (step A7 in FIG. 8). In step A6, if the sequence number is the first (for example, 001), the data storage processing unit 315 does not register the IP address of the data distribution server 2 of the transmission source in the management table 33, and thus the transmission source area The IP address of the data distribution server 2 is set in the previous sequence number area, and the sequence number (eg, 001) of the received packet is set in the previous sequence number area. When the packet is stored in the packet storage unit 34, for example, in the case of the packet having the sequence number 301, the data receiving clients 3-1-1 and 3-1 of the group 1
-2, 3-1-3, 3-1-4, 3-1-5 stores the received packet in the packet storage unit 34, and the sequence number 3
In the No. 02 packet, the group 2 data receiving clients 3-2-1, 3-2-2, 3-2-3, 3-2-
The packets 4 and 2-5 store the received packet in the packet storage unit 34. When the control unit 31 of the data reception client 3 receives the lost packet resend request at the reception unit 311 via the interface unit 32, the control unit 31 causes the reception packet type determination unit 312 to determine whether the UDP segment of the reception packet is included. It is determined from the packet type storage area that the packet has been retransmitted (step A1 in FIG. 8), and the packet retransmit processor 318 determines that the last digit of the sequence number of the packet requested from the UDP segment in the received packet is If the packet belongs to the own group number of the management table 33 (step B1 in FIG. 8) and the requested missing packet is stored in the packet storage unit 34 (step B2 in FIG. 8) (in this case, the management table). The same as the previous sequence number from the data distribution server stored in 33 and the requested sequence number. Then, if it does not match, it is saved. If it does not match, the packet saved for the requesting data receiving client 3 is read from the packet saving unit 34 and the lost packet is retransmitted (packet type storage in FIG. 7). An identification code indicating retransmission of a lost packet in the area is transmitted as the packets read out in the data storage area (step B3 in FIG. 8), and the received request for retransmission of a lost packet is discarded (step B4 in FIG. 5). . That is, the data receiving client 3 transmits the missing packet to the request source in place of the data distribution server 2.

In step B1 or step B2, the control unit 31 causes the missing packet retransmission processing unit 318 to determine that the last digit of the sequence number requested by the missing packet retransmission request does not belong to its own group number. , If the group numbers match but the corresponding packet was not saved, the lost packet resend request is ignored,
Discard (step B4 in FIG. 8). For example, when the data receiving client 3-2-3 detects the packet loss of the sequence number 301, it transmits a lost packet resending request of the sequence number 301 to all the data receiving clients 3 by UDP / IP broadcast communication. Data reception client belonging to group 1 3-
1-1,3-1-2,3-1-3,3-1-4,3-1
When -5 receives this, if the packet of sequence number 301 is stored, the corresponding packet is transmitted to the data receiving client 3-2-3 as a missing packet retransmission. The data receiving clients 3 other than the group 1 ignore the received missing packet retransmission request and discard it. The control unit 31 of the data receiving client 3 that has made a request to retransmit a lost packet is the interface unit 3
When the missing packet retransmission is received by the receiving unit 311 via the H. 2, the control unit 31 determines that the received packet type determination unit 312
At the packet type storage area in the UDP segment of the received packet, it is determined that the missing packet is retransmitted (steps A1 and A2 in FIG. 8), and the missing packet receiving unit 317
Whether or not the missing packet retransmission has already been received is determined by whether or not the area requesting the missing packet retransmission in the management table 33 is "ON" (step C1 in FIG. 8). The reason why the missing packet retransmission is already received is because the missing packet retransmission may be sent from all the data receiving clients belonging to the group storing the corresponding packet.
This is because it is necessary to receive only the packet and discard the lost packet retransmission that is received later.

In step C1, the missing packet retransmission has already been received (the missing packet retransmission request is "OF").
F ”), the control unit 31 causes the correspondence processing unit 316 to discard the packet because it is no longer necessary (step B4 in FIG. 8).

In step C1, if the missing packet retransmission is not yet received (missing packet retransmission request is “ON”), the control unit 31 causes the handling processing unit 316 to delete the missing packet in the management table 33. The requested area is cleared (“OFF”), and processing is performed according to the data content of the received missing packet retransmission (step A7 in FIG. 8). For example, referring to FIG. 9, the data receiving client 3-2-3 detects the packet loss of the sequence number 301, transmits a lost packet resend request, and receives the lost packet resend request.
-1-1,3-1-2,3-1-3,3-1-4,3-
When 1-5 sends the missing packet retransmission to the data receiving client 3-2-3 in this order, the data receiving client 3-2-3 receives the missing packet from the data receiving client 3-1-1 which is first received. Adopt resend,
Data receiving clients 3-1-2, 3 to be received later
The missing packet retransmissions from −1-3, 3-1-4, 3-1-5 are discarded.

Next, a second embodiment of the present invention will be described with reference to the drawings.

The second embodiment of the present invention is the same as the system configuration diagram of FIG. 1 described in the first embodiment, and the internal configuration and operation of the data receiving client 3 are the same as those of the first embodiment. This is different from the embodiment.

That is, referring to FIG. 10, the data reception client 3 is an information processing device such as a workstation or a personal computer, and is controlled by a processor (CPU) under program control and a LAN 1 via the control unit 35. From the data distribution server 2, an interface unit 32 for exchanging data with other devices, a management table 33 for managing own group number, own IP address, previous sequence number of received packet, etc. in connection with UDP / IP packet communication. A packet storage unit 34 that stores the received packet when the last digit of the sequence number of the received packet matches the group number of itself, and a client information table 36 that manages the address information (IP address) of each data receiving client. Consists of.

The management table 33, the packet storage unit 34, and the client information table 36 shown in FIG. 10 are assigned to a certain area in a readable / writable memory (eg, RAM).

Referring to FIG. 11, it is a block diagram showing an example of the configuration in the client information table 36 of FIG. 10, in which the address information (IP address) of each data receiving client 3 is stored in advance for each group. There is. In this case, the client information table 36 has a nonvolatile memory (not shown) built in the data receiving client 3 from an input device (for example, a keyboard) in advance by a maintenance person or the like like the own group number and the own IP address. It is set in an area in a recording medium device (for example, a magnetic disk), and when the data receiving client 3 is started up, the recording medium device is set in the data receiving client 3 in the same manner as the own group number and the own IP address. The client information table 36 is copied to the memory area. In this case, the all data receiving client 3
It has the same client information table 36. It should be noted that the data receiving server 3 is stored in advance, and when the data receiving client 3 starts up, the client information table 36 is automatically sent from the data receiving server 3 so that the data receiving client 3 can store the data. It may be written in the memory.

The data distribution server 2 of FIG. 1 and the management table 33, interface unit 32, and packet storage unit 34 of FIG. 3 have the same configurations as those of the first embodiment, and therefore their explanations are omitted.

Next, with reference to FIGS. 1, 7, and 10 to 12, the operation of the second embodiment will be described centering on the flowchart of FIG. Note that FIG. 12 is a flowchart showing the operation of the control unit 35 when each data receiving client 3 receives a UDP / IP packet.

Now, the data distribution server 2 tries to send the data in the storage unit 23 to the all data receiving clients 3 in order by dividing the data several times. Then, the data distribution server 2 receives the UD as described in the first embodiment.
In P / IP broadcast communication, the UDP segment in FIG. 7 is given a packet type and a sequence number, and information to be delivered is added to perform data delivery. The packet type distinguishes whether the corresponding packet is a delivery packet from a data delivery server, a missing packet resend request requesting resending of a missing packet, or a missing packet resend sent by a missing packet resend request. Is stored in this case, and in this case, it is a delivery packet as the identification code. In addition, the data distribution server 2
Increments the sequence number in the packet from 0 at the time of data distribution.

Control unit 35 of each data receiving client 3
When the distribution packet is received via the interface unit 32, the type determination of the received packet is performed (steps D0, D1, D2 in FIG. 12), and the packet distributed by the data distribution server (the packet type is the distribution server) is determined. If there is, check whether there is any omission in the sequence number received this time, and compare the sequence number of the received packet with the previous sequence number stored in the management table 33 (corresponding to the data distribution server 2 that is the transmission source). (Step D3 in FIG. 12). In step D3, the control unit 35 determines that the first sequence number (for example, 0
If it is 01), it is judged that the sequence number is normal because the transmission source management server 2 is not registered in the management table 33.

The control unit 31 of the data receiving client 3 which detects a missing sequence number (a value obtained by increasing the current sequence number by +2 or more) is stored in the management table 33 (the management table 33). The group number in which the packet is stored is obtained from the last digit of (sequence number) (step D7 in FIG. 12).

Further, the control unit 35 searches the IP address in the client information table 36 based on the detected group number and the self IP address stored in the management table 33 (step D8 in FIG. 12), If there is a matching group number and IP address (if it belongs to the same group as the data receiving client 3 storing the corresponding packet), the IP address registered next to the matching address is matched. If there is no group number and IP address to be used (if it does not belong to the same group as the data receiving client 3 storing the corresponding packet), the data receiving client 3 registered at the head of the matching group Extracts the IP address (in this case, the next IP address if the own IP address is at the beginning) And, the data receiving client 3 missing sequence number is stored the detected packet
Is acquired as the IP address (step D in FIG. 12).
9).

Then, the lost packet resend request is transmitted only to the data receiving client 3 having the IP address detected in step D11 (step D1 in FIG. 12).
0). In this case, the control unit 35 uses the IP address that detected the destination in the packet, the source as its own IP address,
Each is set, and further, the type code of the missing packet retransmission request is added to the packet type area of FIG. 7, the sequence number requested to the sequence number storage area is added, and the IP header of the received packet is added to the data storage area. The management table 33 is transmitted to the all data reception client 3 via the interface unit 32 and
Turn on the area for requesting retransmission of lost packets in the area.

In the processing after step D3 or step D10, the control unit 35 of each data receiving client 3
If the value of the last digit of the sequence number of the received packet belongs to the own group number stored in the management table 33 (step D4 in FIG. 12), the received packet is stored in the packet storage unit 34 (FIG. 12 and step D5), the previous sequence number stored in the management table 33 is replaced with the received sequence number (step D6 in FIG. 12), and processing is performed according to the data content of the received packet (step D7 in FIG. 12). . In step D6, if the sequence number is first (for example, 001), the control unit 35 does not register the IP address of the data distribution server 2 of the transmission source in the management table 33, and thus the data is stored in the transmission source area. The IP address of the distribution server 2 is set in the previous sequence number area, and the sequence number (for example, 001) of the received packet is set. Further, in step D6, if the sequence number is the last, the control unit 35 deletes the information (transmission source, previous sequence number, retransmission request in progress) corresponding to the transmission server 2 of the transmission source from the management table 33.

In step D4, if the last digit of the sequence number of the received packet does not belong to the own group number, the received packet is not saved and the previous sequence number stored in the management table 33 is received. Replace with the packet sequence number (step D in FIG. 12)
6), the process according to the data content of the received packet is performed (step D7 in FIG. 12). Note that in step D6, the control unit 35 determines that the sequence number is first (for example, 0
If it is 01), the IP address of the data distribution server 2 of the transmission source is not registered in the management table 33, so the IP address of the data distribution server 2 is set in the transmission source area, and the sequence number of the received packet is set in the previous sequence number area. (For example, 001) is set.

Control unit 35 of each data receiving client 3
When a missing packet retransmission request is received via the interface unit 32, the corresponding packet is stored in the packet storage unit 34.
(In this case, based on the UDP segment in the received packet, the UD in the received packet from the data delivery server stored in the packet storage unit 34).
If the sequence number of the P segment and the sequence number for which the resend request is made match, it is saved, and if it does not match, it is not saved) and the request source (the sender from the IP header in the UDP segment data storage area) Judgment)
The packet (corresponding to the lost packet) stored in the data receiving client 3 is read out from the packet storage unit 34 and the lost packet is retransmitted (the identification code indicating the lost packet retransmission is set in the packet type storage area of FIG. The packets read out in step S1 are transmitted as settings (steps E1 and E2 in FIG. 12), and the received missing packet retransmission request is discarded (step E in FIG. 12).
6). That is, the data receiving client 3 transmits the missing packet to the request source in place of the data distribution server 2.

In step E1, if the packet is not stored in the packet storage unit 34, the control unit 3
5 refers to its own client information table 36 based on its own address information (step E3 in FIG. 12), and the address information of the data receiving client 3 registered next to itself (in this case, the last in the group). If so, return to the head of the group) (step E in FIG. 12).
4) Transfer the lost packet retransmission request (step E5 in FIG. 12). After that, the data receiving client to which the lost packet retransmission request is transferred also performs the same operation.

For example, referring to FIG. 14, when the data receiving client 3-1-3 detects the packet loss of the sequence number 402, it owns from the last digit of the sequence number 402 where the packet loss occurred. By referring to the area of the group 2 of the client information table 36 that is being operated, the address information of the data receiving client 3-2-1 is acquired as the transmission destination of the missing packet retransmission request. The data reception client 3-1-3 sends the sequence number 402 toward the data reception client 3-2-1.
The data receiving client 3-2-1 which has transmitted the lost packet resending request and has received the missing packet resending request,
If the packet with the sequence number 402 is stored, the packet with the sequence number 402 is transmitted to the data receiving client 3-1-3 as the missing packet retransmission to the data receiving client 3-1-3. Data receiving client 3-2-1 has sequence number 402
If the No. packet has not been stored, the data reception client 3-2-1 refers to its own client information table 36 and refers to the data reception client 3-2-2 registered next to itself in the same group. 2 to transfer the missing packet resend request. Data receiving client 3-
When the 2-2 receives the lost packet resend request with the sequence number 402 transmitted from the data receiving client 3-2-1, if the packet with the sequence number 402 is stored, it sends the lost packet resend request. The packet with the sequence number 402 is transmitted as a missing packet retransmission to the original data receiving client 3-1-3. Further, if the data receiving client 3-2-2 also does not store the packet with the sequence number 402, the data receiving client 3-2-2 refers to its own client information table 36, Data reception client 3 registered next to
-2-3 forwards the missing packet resend request.

When the control unit 31 of the data receiving client 3 which has made the lost packet resend request receives the lost packet resend at the receiving unit 311 via the interface unit 32, the control unit 31 determines that the UDP segment in the received packet is included. It is judged from the packet type storage area that the lost packet is retransmitted (steps D0, D1, and D2 in FIG. 8), the area in the management table 33 in which a retransmission request is made is set to “OFF”, and the data content of the received packet is determined. Perform processing (Fig. 1
2 step D7).

As described in the first and second embodiments above, according to the present invention, in the UDP / IP broadcast communication, the data reception client 3 side receives the UDP / IP data.
When an IP packet loss occurs, the data delivery server 2 is not requested to complement the lost packet, but is requested to the other clients 3, so that it is possible to avoid burdening the data delivery server 2 side.

Further, according to the present invention, the data receiving clients 3 are divided into groups and complement each other for the missing UDP / IP packets. Therefore, when all the received packets are stored, the load is increased and resources are not saved. It is possible to reduce waste.

[0058]

As described above, according to the present invention, the UDP
When the reception UDP / IP packet loss occurs on the data receiving client 3 side in the / IP broadcast communication, the data delivery server does not request the data delivery server to complement the missing packet, so that the data delivery server requests the other client. This has the effect of eliminating the burden on the side.

Further, according to the present invention, the data receiving clients are divided into groups and complement each other for the lost UDP / IP packets. Therefore, when all the received packets are stored, the load is increased and resources are wasted. There is an effect that it is possible to reduce the occurrence of

Further, according to the present invention, since the data reception clients are divided into groups and complement the lost UDP / IP packet, the probability that the packet can be stored increases as the number of data reception clients increases. Therefore, there is an effect that the packet loss can be complemented more reliably.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing a configuration inside the data distribution server of FIG.

FIG. 3 is a block diagram showing an internal configuration of the data reception client of FIG. 1 according to the first embodiment.

FIG. 4 is a block diagram showing a configuration of processing of a program executed by the control unit of the data distribution server of FIG. 1 when transmitting distribution data to a data receiving client.

5 is a block diagram showing a configuration of processing of a program executed by the control unit of the data reception client of FIG. 1 according to the first embodiment when it receives a UDP / IP packet.

6 is a configuration diagram showing a configuration in a management table of the data reception client of FIGS. 3 and 9. FIG.

FIG. 7 is a configuration diagram showing a format of a UDP segment in a UDP / IP packet.

FIG. 8 is a diagram illustrating a case where the control unit of the data distribution server of FIG.
It is a flow chart which shows operation when a P packet is received. The control unit of the data distribution server in FIG.
7 is a flowchart showing an operation when a P / IP packet is received.

FIG. 9 is a schematic diagram showing a packet flow with each data receiving client when the data receiving client 3-2-3 of FIG. 1 according to the first embodiment detects a packet loss.

FIG. 10 is a block diagram showing the internal configuration of the data receiving client of FIG. 1 according to the second embodiment.

11 is a configuration diagram showing a configuration in a client information table of the data reception client of FIG.

12 is a diagram illustrating a case where the control unit of the data distribution server of FIG. 10 uses UDP.
6 is a flowchart showing an operation when a / IP packet is received.

FIG. 13 is a schematic diagram showing a packet flow with each data receiving client when the data receiving client 3-1-3 of FIG. 1 according to the second embodiment detects a packet loss.

FIG. 14 is a block diagram showing a system configuration of a conventional technique.

[Explanation of symbols]

1,4 LAN 2,5 Data distribution server 3,6 Data receiving client 21, 31, 35 Control unit 22, 32 Interface section 23 Memory 33 management table 34 Packet Storage 36 Client Information Table

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Satoshi Terasawa             5-7 Shiba 5-1, Minato-ku, Tokyo NEC Corporation             Inside the company (72) Inventor Tomohiro Suzuki             1-1, Shinurashima-cho, Kanagawa-ku, Yokohama-shi, Kanagawa             Address 25 NEC Robot Engineer Lin             Within Gu Co., Ltd. F term (reference) 5K030 HA08 HC14 JT02 KA01 KA02                       LD02 MB13

Claims (17)

[Claims] 1. A server uses a packet as a UDP to deliver data to a plurality of clients via a network.
In the method of complementing a lost packet in a broadcast communication for transmitting in sequence by / IP broadcast, if the client detects a lost packet while receiving the packet from the server, a request for the lost packet to another client. Broadcast communication characterized by: requesting means for performing the request, and transmitting means for transmitting the missing packet to the request source instead of the server if the other client has stored the requested missing packet. Compensation method for lost packets in. 2. A storage unit that divides the packets received by the clients divided into groups into sequences and divides the packets for each group and stores them in sequence number units, and the requesting unit stores the lost packets. Detected client is UD to other client
The missing packet complement method in broadcast communication according to claim 1, wherein the missing packet is requested by P / IP broadcast. 3. Storage means for sharing the packets received in sequence by the clients divided into groups for each group and storing the packets in sequence number units, and for each group based on the storage means by the client. Management means for managing the destination information of
2. The requesting means further comprises: wherein the requesting means selects a client to be requested from the table by the client which has detected the lost packet, and requests the selected client for the lost packet. Complementary method for missing packets in the described broadcast communication. 4. The server uses a packet as a distribution data for UDP to a plurality of clients via a network.
In a complementary system of a lost packet in a broadcast communication for transmitting in sequence by / IP broadcast, the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a giving means for giving a group number to the clients divided into respective groups The client includes a determination unit that determines the type of the received packet, and, if the determination unit determines that the data is the distribution data from the server, the sequence number of the packet received this time and the previous packet from the server. Detecting means for detecting a missing packet by checking for missing sequence numbers by comparing with the sequence number of the received packet, and a resending request for the missing packet when a missing sequence number is detected by the detecting means. The network Retransmission request means for all the clients connected to, and if the sequence number of the received packet does not detect a missing sequence number in the detection means, the received packet If the requesting sequence number belongs to its own group number when it is judged by the saving means and the judging means that the request for resending the lost packet is made, the requested resending of the missing packet is performed on behalf of the server. A method of complementing a lost packet in broadcast communication, comprising: 5. The server uses a packet as a UDP to deliver data to a plurality of clients via a network.
In a complementary system of a lost packet in a broadcast communication for transmitting in sequence by / IP broadcast, the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a giving means for giving a group number to the clients divided into respective groups The client includes a determination unit that determines the type of the received packet, and, if the determination unit determines that the data is the distribution data from the server, the sequence number of the packet received this time and the previous packet from the server. Detecting means for detecting missing packets by checking for missing sequence numbers by comparing with the sequence number of the received packet, and the missing sequence when the detecting means detects missing sequence numbers The number corresponding to the number Group for extracting a group number, a first selecting section for selecting one of the clients belonging to the extracted group number, and a client for selecting a retransmission request for the lost packet by the first selecting section. And a storage unit for storing the received packet if the sequence number of the received packet belongs to its own group number, if the detection unit does not detect a missing sequence number. And a resending unit that resends the requested lost packet on behalf of the server if the requesting loss packet is stored when the determination unit determines that the lost packet is a resend request. A method of complementing a lost packet in broadcast communication, characterized by including. 6. When the determination unit determines that the request for retransmission of the lost packet is made, one of the clients belonging to the same group as the own group number is selected unless the requested lost packet is saved. Selection method of
6. The method of complementing a lost packet in broadcast communication according to claim 5, further comprising: second retransmission requesting means for making a request to retransmit the lost packet to the client selected by the second selecting means. 7. The method according to claim 4, further comprising: a capturing unit that captures the lost packet if the determination unit determines that the lost packet is retransmitted and the lost packet has not been received. Compensation method for lost packets in broadcast communication. 8. The server uses a packet to send distribution data to a plurality of clients via a network by UDP.
/ IP is a method of complementing a lost packet in broadcast communication transmitted in sequence by the broadcast, wherein the lost packet is sent to another client when the client detects the lost packet while receiving the packet from the server. Requesting the other packet and transmitting the missing packet to the requesting source instead of the server if the other client has stored the requested missing packet. 9. The packets received in sequence by the clients divided into groups are shared for each group and divided and stored in sequence number units, and the client detecting the missing packets requests the missing packets. U to the other client when performing
9. The method of complementing a missing packet in broadcast communication according to claim 8, wherein the missing packet is requested by DP / IP broadcast. 10. The destination information of each group based on the storing means, wherein the packets divided by the client and received in sequence by the clients are shared by the groups and divided and stored in sequence number units. Is managed in a table, and when the client that detects the lost packet requests the lost packet, the client that should request is selected from the table and the lost packet is requested to the selected client. The method of complementing a lost packet in broadcast communication according to claim 8. 11. A server UDs delivery data in packets to a plurality of clients via a network.
A method for complementing a lost packet in broadcast communication for transmitting in sequence by P / IP broadcast, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is assigned to each of the clients divided into respective groups. When the client receives the packet from the server, the client checks whether there is a missing sequence number, and if a missing sequence number is detected, the client sends a request to retransmit the missing packet to the network. If the sequence number sequence is normal and the sequence number of the received packet belongs to its own group number, the received packet is saved, and a request is made when the lost packet is received. of Complement method of lost packets in a broadcast communication, characterized in that Tsu sequence number is transmitted on behalf of the server the lost packet for which the request belongs to the own group number to the requester. 12. A server UDs delivery data in packets to a plurality of clients via a network.
A method for complementing a lost packet in broadcast communication for transmitting in sequence by P / IP broadcast, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is assigned to each of the clients divided into respective groups. When the packet is received from the server, the client checks whether there is a missing sequence number, and if a missing sequence number is detected, the client extracts the group number corresponding to the missing sequence number. Then, one of the clients belonging to the extracted group number is selected, a request for resending a lost packet is made to the selected client, and if the sequence number is normal, the sequence number of the received packet is the own group. Belongs to a number When the save the received packet if that was receiving a retransmission request for the lost packet,
A method of complementing a lost packet in broadcast communication, comprising transmitting the requested lost packet to a request source instead of the server if the requested lost packet is stored. 13. The client, when receiving the lost packet, fetches the lost packet if not already received, and discards the lost packet if already received. Alternatively, the method of complementing a lost packet in broadcast communication according to item 12. 14. A server UDs delivery data in packets to a plurality of clients via a network.
A method for complementing a lost packet in broadcast communication for transmitting in sequence by P / IP broadcast, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is assigned to each of the clients divided into respective groups. When the client determines the type of the received packet and determines the type of the received packet as distribution data from the server, the client determines the sequence number of the packet received this time and the server last time. If a missing packet is detected by checking if there is a missing sequence number by comparing with the sequence number of the packet received from Before resend request Performed for all the clients connected to the network, and if the sequence number is normal by checking whether the sequence number is complete, the received packet if the sequence number of the received packet belongs to its own group number Stored, and after checking whether the sequence number of the received packet belongs to its own group, the packet sequence number received this time is replaced with the sequence number of the previously received packet, and the type of the received packet is When it is determined that the request is a retransmission request, if the requested sequence number belongs to its own group number, the requested lost packet is retransmitted on behalf of the server. Method. 15. The server UDs delivery data in packets to a plurality of clients via a network.
A method for complementing a lost packet in broadcast communication for transmitting in sequence by P / IP broadcast, wherein the plurality of clients are divided into groups corresponding to the sequence numbers of the packets, and a group number is assigned to each of the clients divided into respective groups. When the client determines the type of the received packet and determines the type of the received packet as distribution data from the server, the client determines the sequence number of the packet received this time and the server last time. The sequence number of the packet received from is checked to see if there is a missing sequence number and a missing packet is detected. If the missing sequence number is detected by checking if there is a missing sequence number, the missing Matches the sequence number Group number is extracted, one of the clients belonging to the extracted group number is selected, a request for resending a lost packet is made to the selected client, and the sequence is checked by checking whether the sequence number is complete. If the number is normal, the received packet is saved when the sequence number of the received packet belongs to its own group number, and the packet received this time after checking whether the sequence number of the received packet belongs to the own group When the sequence number is replaced with the sequence number of the previously received packet, and the type of the received packet is determined to be a retransmission request for the lost packet, the requested lost packet is stored and the requested packet is stored. Special feature is to resend lost packets on behalf of the server. Complement method of lost packets in a broadcast communication to. 16. If one of the clients belonging to the same group number as the own group number is selected if the type of the received packet is determined to be a retransmission request for the lost packet and the requested lost packet is not saved. 16. The method of complementing a lost packet in broadcast communication according to claim 15, wherein the request for retransmitting the lost packet is made to the client selected by the second selecting means. 17. When the type of the received packet is determined to be the retransmission of a lost packet, the lost packet is fetched if the lost packet has not been received,
16. The method for complementing a lost packet in broadcast communication according to claim 15, wherein the lost packet is discarded if the lost packet has already been received.