JP2000253061A - Data delivery method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data delivery method where processing capability consumed by a server for reservation of a resource amount is almost not increased even when the number of reception nodes being simultaneous delivery destinations is increased. SOLUTION: Information to identify a transfer rate and a reception node is sent/received between a transmission node 101, reception nodes 104, 106, routers 103, 105 belonging to the same network. The routers 102, 103, 105 receiving this identification information execute resource reservation and construction of routing information. Then the transmission node 101 transmits consecutive media data by a packet in compliance with an IP compatible format. The routers 102, 103, 105 copy packets through DMA without adopting memory copy. The routers 103, 105 belonging to the same network as the reception nodes 104, 106 convert the received packet into an IP packet and delivers the IP packet to the reception nodes.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data distribution method, and more particularly, to a method for transmitting continuous media data such as audio data and moving image data from an information transmitting device to a plurality of information processing devices via a plurality of information relay devices. The present invention relates to a data delivery method suitable for executing simultaneous delivery (multicast) of the data.

[0002]

2. Description of the Related Art An information transmission apparatus (hereinafter, referred to as a "transmission node") transmits a plurality of information relay apparatuses (hereinafter, referred to as "routers").
) Connected through a plurality of nodes (hereinafter, referred to as
As a method of simultaneously delivering continuous media data such as audio data and video data to the "receiving node", RTIP (RealTime Internet Protocol) and RTIPSI
Methods using G (RealTime Internet Protocol SIGnaling Protocol) are described, for example, in "Osamu Takeuchi et al.," Design and Implementation of QoS Guaranteed Routing Method Applying Isochronous Scheduler ", Information Processing Society of Japan, Multimedia Communication and Distributed Processing Workshop Transactions, (1998 11
Mon), pp 119-126 ".

For simultaneous delivery of continuous media data, a reservation process for reserving various resources and a delivery process for continuous media data are required in advance.

In the method using RTIP and RTIPSIG, the transmitting node allocates resources (such as router CPU time, buffers, and network bandwidth) required for data transfer before starting transmission of continuous media data. To secure
The CONNECT message and the ACCEPT message are exchanged between the transmitting node and the router adjacent to each receiving node. Here, the adjacent router refers to a router belonging to the same Ethernet network as the receiving node.

[0005] The CONNECT message and the ACCEPT message store information on the transfer rate of continuous media data to be transmitted later. Each router that has received the ACCEPT message reads the transfer rate information stored in the ACCEPT message, and secures resources necessary for relay processing of continuous media data arriving at the corresponding transfer rate. At this time, the transmitting node transmits a CONNECT message and A
Since the transmission and reception of CCEPT messages must be repeated as many times as the number of receiving nodes, the processing overhead required for reservation processing before the start of continuous media data delivery increases in proportion to the number of receiving nodes, and one transmission There has been a problem that the number of receiving nodes that can simultaneously deliver continuous media data cannot be increased.

[0006] When the transmission and reception of the CONNECT message and the ACCEPT message are completed, the transmitting node starts transfer of continuous media data by transmitting an IP compatible packet to the receiving node. The packet format of the IP compatible packet is the same as the IP packet, and can be received by the existing IP protocol stack. Conventional IP
The difference from the packet is that each router that executes the relaying process of the packet uses the resources secured when receiving the ACCEPT message without excess or shortage, and relays the packet at a rate exactly equal to the packet arrival rate. The point is to execute the processing. Since each router performs packet relay processing at a rate exactly equal to the arrival rate, it is guaranteed that the arrival rate of continuous media data at the receiving node exactly matches the transmission rate of continuous media data at the sending node. it can. At this time, since the transmitting node needs to repeat the transfer of the continuous media data by the number of the receiving nodes, the processing overhead required for the delivery process of the continuous media data in the transmitting node increases in proportion to the number of the receiving nodes. However, there is a problem that the number of receiving nodes to which a single transmitting node can simultaneously deliver continuous media data cannot be increased.

On the other hand, in the method using ST-II (RFC-1170), before starting transmission of continuous media data, the transmitting node secures resources necessary for transferring the data. For this purpose, a CONNECT message is transmitted to all receiving nodes. The CONNECT message stores IP address information of the receiving node group and information on the transfer rate of continuous media data to be transmitted later.

[0008] The router that has received the CONNECT message, C
The transfer rate information stored in the ONNECT message is read, and resources necessary for the relay processing of the continuous media data arriving at the corresponding transfer rate are secured. In addition, CONNEC
The IP address information of the receiving node group stored in the T message is read, and the CONNECT message is transferred to one or more routers or receiving nodes so that the CONNECT message reaches all the receiving nodes.

Each receiving node that has received the CONNECT message transmits an ACCEPT message to the transmitting node. When the transmitting node receives the ACCEPT message from all receiving nodes, it starts transferring continuous media data.

[0010] Simultaneous delivery of data is realized by storing continuous media data in an ST packet and delivering the ST packet to each receiving node. In the simultaneous delivery using ST packets, even when continuous media data is delivered to a plurality of receiving nodes, the sending node does not need to send a plurality of ST packets storing the same data. However, ST
The router that has received the packet transmits the ST packet to a plurality of routers so that the ST packet reaches all the receiving nodes. Therefore, each time the ST packet is received, the packet is copied (memory copy). .

[0011]

[Problems to be solved by the invention] However, ST-II
In the method using, as described above, since the transmitting node needs to receive the ACCEPT message from all receiving nodes, the overhead required during resource amount reservation processing,
There is a first problem that the number increases in proportion to the number of receiving nodes.

In addition, since the router needs to generate copies of the continuous media data (perform memory copy) by the number of networks to be transmitted, the processing capacity of the router is saturated by the memory copy overhead, and the receiving node has to be copied. There is a second problem that simultaneous delivery cannot be realized.

A first object of the present invention is to provide a data delivery method in which the processing capacity consumed by a server for resource amount reservation processing hardly increases even if the number of receiving nodes serving as simultaneous delivery destinations increases. It is in.

A second object of the present invention is to provide a data processing system in which the processing capacity consumed by a router for transmitting data to a network group hardly increases even if the number of networks to which continuous media data is to be transmitted simultaneously increases. To provide a delivery method.

[0015]

Means for Solving the Problems (1) In order to achieve the first object, according to the present invention, an information transmitting apparatus is provided to a plurality of information receiving apparatuses via a plurality of information relay apparatuses.
In the data delivery method for simultaneously delivering data after performing resource reservation processing, A) a request for data reception is made between the information transmitting apparatus and an information relay apparatus belonging to the same network as the plurality of information receiving apparatuses. Transmitting and receiving identification information for identifying the plurality of information receiving apparatuses and control information relating to a data transfer rate requested by the information transmitting apparatus; and B) the information relay apparatus receiving the information transmitting apparatus and the control information. A device for reserving resources required for data transfer; and C) the information transmitting device and the information relay device receiving the control information, the information relay device to be a next relay destination of data or Constructing identification information for identifying the information receiving device;
D) At the end of the resource reservation, after receiving a resource reservation end message from all of the information relay devices in the next stage, the information relay device or the information transmitting device transmits the resource reservation end to the preceding stage. And transmitting a message to execute resource reservation processing. With this method, even if the number of receiving nodes serving as simultaneous delivery destinations increases, the processing capacity consumed by the server for resource amount reservation processing hardly increases.

(2) To achieve the second object,
The present invention relates to a data delivery method for performing resource reservation processing from an information transmitting apparatus to a plurality of information receiving apparatuses via a plurality of information relay apparatuses, and then simultaneously delivering data. The transmitting device reads the identification information for identifying the information relay device or the information receiving device built in the information relay device by the resource reservation processing,
Transmitting data to the read information relay device or information receiving device in accordance with the requested data transfer rate; and F) the information relay device having received the transmitted data transmits the information relay device or information to be the next relay destination. The identification information for identifying the receiving device is read out, and the transmitted data is transmitted to the network interface group connecting the information relaying device or the information receiving device to be the next relay destination read out using the DMA. And a step of requesting transmission. With this method, even if the number of networks to which continuous media data is to be transmitted simultaneously increases, the processing capacity consumed by the router for transmitting data to the network group hardly increases.

(3) In the above (2), preferably,
G) The information relay device that belongs to the same network as the information receiving device that has received the transmitted data, changes the data format in which the transmitted data is stored only from the information transmitting device to the information receiving device. Converting the data into a data format to be used when performing data transmission. By such a method, for an information receiving device that can only support the existing data format,
Simultaneous delivery of continuous media data can be performed.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A data delivery method according to an embodiment of the present invention will be described below with reference to FIGS. First, the configuration of a data delivery system that implements the data delivery method according to the present embodiment will be described with reference to FIG. FIG. 1 is a system block diagram showing a configuration of a data delivery system for realizing a data delivery method according to one embodiment of the present invention.

The data delivery system according to the present embodiment includes a transmitting node 101 as an information transmitting apparatus, a plurality of receiving nodes 104 and 106 as an information receiving apparatus, and information between the transmitting node 101 and the receiving nodes 104 and 106. Routers 102, 10 which are information relay devices for relaying
3, 105. Between the transmitting node 101 and the router 102, between the router 102 and the router 103,
Between the router 102 and the router 105, between the router 103 and the receiving node 104, and between the router 105 and the receiving node 10.
6 are connected by Ethernet (registered trademark) networks 107, 108, 109, 110, and 111, respectively.

In the above system configuration, the number of receiving nodes is two for the sake of simplicity, but actually two or more receiving nodes are connected via an Ethernet network. The number of routers connected to the transmitting node is not one but plural.

The simultaneous delivery according to the present embodiment is realized by issuing the following requests from an application on the transmission node 101. 1) Issuing a batch resource reservation request (allocate_resource function) 2) Issuing a simultaneous delivery request (rt_multisend function) The batch resource reservation request specifies the IP address of the receiving node group and the transfer rate of continuous media data, and the transmitting node 1
Resources required for simultaneous delivery of continuous media data at the specified rate to all specified receiving nodes from 01 (router buffer, CPU time, network bandwidth)
Is required. The sending node 101 and the router 10 when the batch resource reservation request is issued
Operations of 2, 103, and 105 will be described later with reference to FIGS.

The simultaneous delivery request is for simultaneously delivering continuous media data specified by the transmitting node 101 to all receiving nodes specified when the batch resource reservation request is issued. At this time, the arrival rate of the continuous media data at each receiving node exactly matches the transmission rate of the continuous media data at the transmitting node. When the simultaneous delivery request is issued, the transmitting node 101, the routers 102 and 10
3 and 105 and the operations of the receiving nodes 104 and 106 will be described later with reference to FIGS.

Hereinafter, referring to FIGS. 1 to 9, the transmission node 101 and the router 1 when the batch resource reservation request is issued
Operations of 02, 103 and 105 will be described. First, referring to FIG. 1, the transmitting node 101 according to the present embodiment will be described.
When the above application issues the batch resource reservation request, the transmitting node 101 and each of the routers 102, 103, 1
An outline of the operation of the operation 05 will be described.

In the example shown in FIG. 5, the transmitting node 101
It requests the receiving node 104 and the receiving node 106 to secure resources required for simultaneous delivery of continuous media data. The sending node 101
The ECT message is delivered to routers adjacent to all the receiving nodes 104 and 106. Here, "adjacent router" means
The routers 103 and 105 belonging to the same Ethernet network as the receiving nodes 104 and 106 are called.

First, the transmitting node 101 sends a CONNECT message 120 to the router 102, which is the next-stage relay device.
Send Here, the CONNECT message 120 includes
As described later, the receiving node group (receiving nodes 104, 1
06) IP address information and information on the transfer rate of continuous media data to be transmitted later are stored.

The router 102 that has received the CONNECT message 120 reads the IP address information of the group of receiving nodes stored in the CONNECT message 120 and sends it to all the neighboring routers 103 and 105 of the corresponding receiving node.
The CONNECT messages 121 and 122 are forwarded to one or more routers 103 and 105, respectively, so that the CONNECT message 120 arrives. When another router is interposed between the router 102 and the adjacent router 103 to the receiving node 104, the router 102 transfers a CONNECT message to the intermediate router, and further, the intermediate router transmits a CONNECT message to the adjacent router 103.
By transferring the message, a CONNECT message is finally transferred to the neighboring router 103.

Adjacent routers 103, 1 of receiving nodes 104, 106 that received CONNECT messages 121, 122
In step 05, information on the transfer rate of continuous media data transmitted later is read from the CONNECT messages 121 and 122. Then, resources sufficient to transfer the continuous media data arriving at this transfer rate are reserved. This resource reservation method is based on the method using RTIP and RTIPSIG (Osamu Takeuchi et al., “Design and Implementation of QoS Guaranteed Routing Method Applying Isochronous Scheduler”, Proceedings of Multimedia Communication and Distributed Processing Workshop, p.
p119-126). Further, the adjacent routers 103 and 105 of the receiving nodes 104 and 106
Source nodes of the T messages 121 and 122 (hereinafter referred to as
ACCEPT messages 123 and 124 are returned to the “previous hop node”.

A router 10 that is not a router adjacent to the receiving node
2 is an ACCEPT message 123, 12 from all the routers 103, 105 that have transmitted the CONNECT messages 121, 122 (hereinafter, the destination nodes of the CONNECT messages 121, 122 are referred to as "next hop nodes").
Wait until 4 is received. When ACCEPT messages 123 and 124 are received from routers 103 and 105, which are all next hop nodes, ACCEPT messages are sent to transmission node 1010 which is the previous hop node of the own node.
Forward the message 125. The sending node 101
When the CEPT message 125 arrives, the corresponding node recognizes that the batch resource reservation request has been completed. That is, in the present embodiment, the next hop node, upon receiving all the ACCEPT messages,
Since the PT message is transmitted, the transmitting node 101, which is the preceding hop node, transmits the PT message from all the receiving nodes.
This eliminates the need to receive the ACCEPT message and prevents the overhead required for resource amount reservation processing from increasing in proportion to the number of receiving nodes.

Next, referring to FIG. 2, the ACCEPT shown in FIG.
The module configuration of the transmission node 101 and the routers 102, 103, and 105 for transmitting and receiving a message and a CONNECT message will be described. FIG. 2 is a block diagram illustrating a module configuration of a transmission node and a router used in the data delivery method according to the present embodiment.

Transmission node 101 and routers 02 and 10
3, 105, the input side network interface 60
2 is connected to the input side network 601. However, regarding the transmitting node 101, the input side network interface 602 does not necessarily need to be provided. The transmitting node 101 and the routers 02, 103, and 105 are connected to an output network 609 through an output network interface 608.

The network interface receiving module 603 receives a packet arriving through the input side network 601 and the input side network interface 602, and
4 is started and this packet is transferred.

The IP protocol stack module 604
Executes the IP protocol processing on the received packet. That is, the IP protocol stack module 604
Refers to the IP header part of the received packet, and
Activation of the IG protocol stack module 605 and RTIPSIG protocol stack module 60
5 and relay processing of a packet for which a transmission request has been issued from the RTIPSIG protocol stack module 605. For the IP protocol processing and the packet relay processing, for example, "
"Open Design No.3-Ethernet (registered trademark) TCP"
/ IP- ", CQ Publishing Company, pp. 28-32".

The RTIPSIG-multi protocol stack module 605 executes an RTIPSIG-multi protocol process on a received packet. In the RTIPSIG-multi protocol processing, a resource used for simultaneous delivery of IP compatible packets to be executed later is reserved, and an RTIP routing table 610 is constructed. RTIPSIG-multi protocol stack module 605, if necessary, RTIPSIG-mult
It issues an i-packet transmission request and notifies the application of the success of resource reservation. The RTIPSIG-multi packet for which the transmission request has been issued is sent to the output side network 60 through the IP protocol stack module 604 and the network interface transmission module 607.
9 is transmitted. In the transmitting node 101,
RTIPSIG-multi protocol stack module 605
May be activated upon issuance of a batch resource reservation request from the application 606.

Here, the data structure used by each module will be described with reference to FIG. FIG. 3 is an explanatory diagram of a data structure used by each module in the data delivery method according to the present embodiment.

FIG. 3 shows CONNECT messages 120, 121, 122 and ACCEPT messages 123, 1 shown in FIG.
24 shows an RTIPSIG-multi packet format which is a format of 24, 125.

The RTIPSIG-multi packet is composed of an IP header 701 and an RTIPSIG-multi header 702. The IP header section 701 includes the IP address of the conventional IP packet.
It has the same format as the P header section. Regarding the IP header part of the conventional IP packet, for example, "[Open Des
ign No. 3 -Ethernet TCP / IP- ", CQ Publishing Company, pp. 28 to 32". In the process of the present embodiment shown in FIG. A number field 703 and a destination address field 704.

The RTIPSIG-multi header 702 has the following fields. Command field 70
5 stores the type of CONNECT or ACCEPT command. The QoS parameter field 706 stores the resource reservation amount requested by the application 606 on the transmission node 101 that has issued the batch resource reservation request. The stream ID field 707 stores a stream ID to be used for an IP compatible packet that is simultaneously delivered later by a simultaneous delivery request. "Stream I
“D” refers to the stream to which the IP compatible packet arriving at the router belongs (the application 6 on the transmitting node 101).
Reference numeral 06 denotes an ID for identifying a transmission node and a reception node group specified when the batch resource reservation request is issued. Each router determines, based on the stream ID, which of the resources reserved in the batch resource reservation request is to be used to execute the relay processing of the corresponding IP compatible packet.
The preceding hop node address field 708, the next hop node address field 709, the transmitting node address field 710, and the receiving node address field groups 711 and 712 respectively include a preceding hop node, a next hop node, the transmitting node 101, and a receiving node group. I
Stores the P address.

Next, referring to FIG. 4, the RTI shown in FIG.
The data structure of the P routing table 610 will be described. FIG. 4 is an explanatory diagram of the data structure of the RTIP routing table used in the data delivery method according to the present embodiment.

The RTIP routing table 610 is
It comprises a stream ID index 801, a routing table entry 802, and a next hop node entry group 811A, 811B,.

The stream ID index 801 is obtained from the stream ID by the routing table entry 802.
It is intended to execute the conversion to a high speed. The stream ID index 801 is an array of pointers to the routing table entry 802. The index value of this array is stored in each routing table entry 80
2 is the same as the stream ID corresponding to 2. Note that the routing table entry 802 exists for each stream.

The next-stage hop node entry group 811
A, 811B,... Are connected in a list from the routing table 802 by pointers. The routing table entry 802 includes the following fields. The stream ID field 803 stores a stream ID corresponding to the corresponding routing table entry 802. Sending node address field 804, receiving node address field group 805
-1,..., 805-n, the preceding hop node address field 809 stores the IP address of the transmitting node 101, the receiving node group, and the preceding hop node belonging to the stream specified by the stream ID. The QoS parameter field 807 stores the QoS parameter specified when the application on the transmission node 101 issues the batch resource reservation request. The host address field 808 stores the IP address of the network interface connected to the network between the own node and the previous hop node. The next-hop node list field 810 stores a pointer to a list of next-hop node entries 811A.

Next hop node entry 811A, 81
1B is composed of the following fields. The next-hop node entries 811A and 811B are:
Each has the same configuration. The next-hop node address fields 812A and 812B store the IP address of the next-hop node. The stream ID fields 813A and 813B store the stream ID assigned to the corresponding stream by the next hop node (even for the same stream, the assigned stream ID differs for each node). The output interface name fields 814A and 814B store an output network name when transmitting an IP compatible packet from the own node to the next hop node. The host address fields 815A and 815B store the IP addresses assigned to the output network interfaces. Receive node address field group 8
16A-1,..., 816A-n, 816B-1,.
16B-n stores the IP address of the receiving node group that can be reached via the next-stage hop node. The continuation bits 818A and 818B store bits indicating whether or not to wait for the arrival of the ACCEPT message.
The local bits 819A and 819B store bits indicating whether the next-hop node is the receiving node.

For example, in the configuration of the data delivery system shown in FIG. 1, the routing table of the transmitting node 101 is configured as follows. That is, when the receiving nodes are two receiving nodes 104 and 106,
The I of the receiving node 104 is added to the receiving node address 805-1.
The P address is stored, and the receiving node address 805-2 is stored.
Stores the IP address of the receiving node 106. Also, as shown in FIG. 1, since one router 102 is connected to the transmission node 101, the next-stage hop node entry 811 is only one next-stage hop node entry 811A, and The node address 812A stores the IP address of the router 102.

On the other hand, for example, in the configuration of the data delivery system shown in FIG. 1, the routing table of the router 102 is configured as follows. That is,
If the receiving nodes are two receiving nodes 104 and 106, the receiving node address 805-1 stores the IP address of the receiving node 104, and the receiving node address 805-2 stores the IP address of the receiving node 106. Is done. However, since the router 103 as a relay device for the receiving node 104 is different from the router 105 as a relay device for the receiving node 106, as shown in FIG.
3 and 105, the next-hop node entry 811 is a next-hop node entry 811A and a next-hop node entry 811B, and the next-hop node address 812A contains the IP address of the router 103.
The next hop node address 812 is stored.
B stores the IP address of the router 105. The receiving node address 816A-1 of the next-hop node entry 811A stores the IP address of the receiving node 104, and the receiving node address 816B-1 of the next-hop node entry 811B stores the receiving node 1016.
6 are stored.

Next, referring to FIGS. 5 to 9, FIG.
Transmission and reception of a CONNECT message in the RTIPSIG-multi protocol stack module 605, and A
The operation of transmitting and receiving a CCEPT message will be described. FIGS. 5 and 6 are flowcharts illustrating the operation of transmitting a CONNECT message in the RTIPSIG-multi protocol stack module in the data delivery method according to the present embodiment. FIG. 7 is a flowchart illustrating the operation of the RTIPSIG-multi protocol in the data delivery method according to the present embodiment. FIG. 8 is a flowchart illustrating an operation of transmitting and receiving a CONNECT message and an operation of transmitting an ACCEPT message in the stack module. FIG. 8 is a flowchart illustrating an operation of receiving a CONNECT message in the RTIPSIG-multi protocol stack module in the data delivery method according to the present embodiment. FIG. 9 is a flowchart illustrating an ACCEPT message receiving operation in the RTIPSIG-multi protocol stack module in the data delivery method according to the present embodiment.

First, the operation when the RTIPSIG-multi protocol stack module 605 in the transmitting node 101 transmits the CONNECT message 120 will be described with reference to FIGS.

The application on the transmitting node 101 issues a batch resource reservation request using the following interface. <Function name> allocate_resource (sender_ip_addr, recver_ip_add
r [], QoS_parameter) <Argument> sedner_ip_addr: Specifies the IP address of the sending node. recver_ip_addr: Specifies the IP address of the receiving node group. QoS_parameter: Specifies a QoS parameter. <Return Value> The stream ID secured in step 901 (described later) returns.

<Explanation> A resource reservation required for transferring an IP compatible packet based on the QoS parameter specified by QoS_parameter is requested between the transmitting node specified by sender_ip_addr and the receiving node specified by recver_ip_addr. When the above function is issued, the application 606 on the sending node 101 sends the RTIPSIG-multi module 605
And pass the arguments of the above interface.

In step s100 of FIG.
G-multi module 605 uses unused stream ID
And secure a routing table entry area.

Next, at step s110, the RTIPSI
The G-multi module 605 checks the stream I in the routing table entry area secured in step s100.
The value of the stream ID secured in step s100 is set in the D field 803. Further, the values of the transmission node address field 804, the reception node address field group 805-a,..., 805-n, and the QoS parameter field 807 are set according to the arguments passed from the application 606. Further, “0” representing its own node is stored in the host address field 808 and the preceding hop node address field 809.

In step s120, RTIPSIG-mult
The i module 605 determines a next-hop node group, an output interface name group, and an IP address group bound to the output interface for the reception node address field groups 805-1,. . Regarding the IP protocol routing solution, for example, see "Open Design
No.3 -Ethernet TCP / IP- ", CQ Publisher,
pp. 28-32 ".

In step s130, RTIPSIG-mult
The i module 605 creates the next-stage hop node entries 811 by the number of next-stage hop node groups obtained in step s120. Then, the next-hop node address field 812, output interface name 814, and host address field 815 of the next-hop node entry are stored with the result obtained by the routing solution in step s120. .., 816-n are stored in the receiving node field groups 816-1,..., 816-n. Further, “0” is stored in the continuation bit field 818. Finally, the local bit 819 is set to "1" or "0" depending on whether the receiving node belongs to the same segment as the own node.

Next, in step s140 of FIG.
The RTIPSIG-multi module 605 performs step s130
All next-hop node entries 8 created in
Steps s150, s160, and s170 for 11
Alternatively, it is searched whether or not steps s150, s165, and s170 have been executed. When the execution of the above step group is completed for all the next-hop node entries 811, the process is completed in step s180. Otherwise, jump to step s150.

At step s150, the local bit 81 of the next hop node entry 811 of interest
It is determined whether 9 is “0”. If "0",
If “1” is set in step s160, step s16
Jump to 5. Note that the transmitting node 101
At the stage of transmitting the T message, since the local bit 819 of the next hop node entry 811 is “0”, the process jumps to step s160.

Next, at step s160, the RTIPSI
The G-multi module 605 generates an RTIPSIG-multi packet (CONNECT message 501). RTI to generate
Protocol number field 703 of PSIG-multi packet
Next, the protocol number of the RTIPSIG-multi protocol is stored in the destination address field 704 in the next hop node address 811 of the next hop node entry 811 of interest.
12 values are stored. In addition, the command field 70
5 stores a value indicating the CONNECT message 501.
Further, the value of the host address field 815 of the next hop node entry 811 focused on the previous hop node address field 708 is stored. Also, Qo
S parameter 706, sender node address field 7
10 and receiving node address field group 711, 71
2 is the next-hop node entry 811 of interest.
Of the routing table entry 802 corresponding to
, An S parameter field 807, a transmission node address field 803, and a reception address field group 816,.
The value stored in 816-n is set.

Next, in step s170, the RTIPSI
The G-multi module 605 activates the IP protocol stack module 604 to transmit the RTIPSIG-multi packet generated in step s160, and
Request transmission of i-packet.

When transmitting the CONNECT message 120, the operations of the network interface receiving module 603, the IP protocol stack module 604, and the network interface transmitting module 607 of the transmitting node 101 are the same as those in the conventional IP protocol stack.

Next, referring to FIG. 8 and FIG.
2 receives the CONNECT message 120 from the transmitting node 101 and the CO
The transmission operation of the NNECT messages 121 and 122 will be described. In step s200 of FIG.
The second RTIPSIG-multi module 605 secures an unused stream ID and a routing table entry area.

In step s210, RTIPSIG-mult
The i module 605 sets the value of the stream ID secured in step s200 in the stream ID field 803 of the routing table entry area secured in step s200. Further, a transmission node address field 804 and a reception node address field group 805
-1, ..., 805-n, QoS parameter field 8
07, the transmitting node address field 710, the receiving node address field group 711, 712, and the QoS parameter field 70 of the CONNECT message 120 that has received the value of the previous hop node address field 809.
6. Set to the value of the previous hop node address field 708. Further, the host address field 808 of the corresponding routing table entry 802 is received.
Destination address field 7 of CONNECT message 120
Set to the value of 04.

In step s220, RTIPSIG-mult
The i module 605 determines a next-hop node group, an output interface name group, and an IP address group bound to the output interface for the reception node address field groups 805-1,. . Regarding the IP protocol routing solution, for example, see "Open Design
No.3 -Ethernet TCP / IP- ", CQ Publisher,
pp. 28-32 ".

In step s230, RTIPSIG-mult
The i module 605 creates the next-stage hop node entries 811 by the number of next-stage hop node groups obtained in step s120. Then, the next-hop node address field 812, output interface name 814, and host address field 815 of the next-hop node entry are stored with the result obtained by the routing solution in step s120. .., 816-n are stored in the receiving node field groups 816-1,..., 816-n. Further, “0” is stored in the continuation bit field 818. Finally, the local bit 819 is set to "1" or "0" depending on whether the receiving node belongs to the same segment as the own node. Next, returning to step s140 of FIG.
0, s170, the next router 10
CONNECT messages 121 and 12 for 3 and 105
Send 2.

Next, referring to FIG. 8, FIG. 6, and FIG.
Receiving operation of T messages 121 and 122 and router 10
The transmission operation of the ACCEPT messages 123 and 124 for No. 2 will be described. The operation of receiving the CONNECT messages 121 and 122 from the router 102 in the routers 103 and 105 is the same as the operation of receiving the CONNECT message 120 of the router 102, and the steps s200 and s2 in FIG.
10, s220 and s230 are sequentially executed.

Next, in step s140 of FIG.
RTIPSIG-multi module 60 for routers 103 and 105
5 corresponds to step s15 for all next-hop node entries 811 created in step s230.
0, s160, s170, or steps s150, s
It is searched whether or not 165 and s170 have been executed. When the execution of the above step group is completed for all the next-hop node entries 811, the process is completed in step s180. Otherwise, step s15
Jump to zero.

In step s150, the local bit 81 of the next hop node entry 811 of interest
It is determined whether 9 is “0”. If "0",
If “1” is set in step s160, step s16
Jump to 5. Since the routers 103 and 105 are adjacent nodes to the receiving nodes 104 and 106, the local bit 819 of the next-stage hop node entry 811 is “1”, so that the process jumps to step s165.

Then, in step s165 of FIG. 7, the RTIPSIG-multi module 605
Generate packets (ACCEPT messages 123 and 124). In step s165, the protocol number field 703 of the RTIPSIG-multi packet contains the RTIPSIG-multi
The protocol number of the protocol is stored in the destination address field 704 in the next hop node entry 8 of interest.
11 is stored in the host address field 815. Then, resource reservation is executed. Further, a value indicating the ACCEPT messages 123 and 124 is stored in the command field 705. Further, the next-hop node address field 812 of the next-hop node entry 811 focused on the next-hop node address field 709
Store the value of. The QoS parameter 706, the transmission node address field 710, and the reception node address field group 711, 712 include the QoS parameter field 8 of the routing table entry 802 corresponding to the next-hop node entry 811 of interest.
07, the transmission node address field 803, and the values stored in the reception address field groups 816-1,..., 816-n of the focused next hop node entry 811 are set. Finally, "0" is stored in the stream ID field 707.

Next, referring to FIG. 9, an ACCEPT message 12 from the routers 103 and 105 in the router 102 will be described.
3,124 receiving operation and AC for transmitting node 101
The transmission operation of the CEPT message 125 will be described.

At step s300, the router 102
RTIPSIG-multi module 605 of received ACCEPT
The transmission node address field 710 and the reception node address field group 711 of the messages 123 and 124
712, the value of the next-hop node address 709 matches all the values of the transmission address field 804 of the routing table entry 802, the receiving node address field group 816, 817 of the next-hop node entry 811 and the next-hop node address field 812. The combination of the routing table entry 802 and the next hop node entry 811 to be performed is searched.

Next, in step s310, the RTIPSI
The G-multi module 605 sets the continuation bit 818 of the next-hop node entry 811 retrieved as a result of step s300 to “1”. Further, the value of the stream ID field 813 of the corresponding next-hop node entry 811 is set in the stream I of the received ACCEPT message 502.
It is set to the value of the D field 707.

Next, at step s320, the RTIPSI
The G-multi module 605 searches whether or not the value of the continuation bit field 818 of all the next-hop node entries 811 connected to the routing table entry 802 searched as a result of step s300 is “1”. . If all are “1”, step 1304
Jump to Otherwise, the process is completed. Here, the retrieved routing table entry 802
Next hop node entry 8 connected to
11 indicates that the value of the continuation bit field 818 is “1”, which means that all routers 1 that transmitted the CONNECT message
This is the case where an ACCEPT message has been sent from 03, 105. In this step s320, the presence or absence of all the ACCEPT messages from the next router is checked.

When the value of the continuation bit field 818 is “1”
, At step s330, the RTIPSIG-multi
The module 605 resets the value of the all continuation bits field 818 retrieved in step s320 to “0”.
Further, it executes resource reservation necessary for transferring the IP compatible packet later. That is, the router 102 is the router 1
When all the ACCEPT messages are received from 03 and 105, resource reservation is made. The method of resource reservation is the same as the method using RTIP and RTIPSIG. RTI
For details on resource reservation methods using P and RTIPSIG, see, for example, "Takashi Takeuchi et al.," Design and Implementation of QoS Guaranteed Routing Method Applying Isochronous Scheduler ", Proceedings of Multimedia Communication and Distributed Processing Workshop, p.
pp. 119-126.

Next, at step s340, the RTIPSI
The G-multi module 605 searches whether or not the value of the previous hop node address field 809 of the routing table entry 802 searched as a result of Step s300 is “0”. If "0", step s35
At 0, the application is notified of the success of the resource reservation and the process is completed. If not "0", step s3
Jump to 60. Here, the previous hop node address field 8 of the routing table entry 802
The case where the value of 09 is “0” is the case of the transmitting node 101, and the case of the router 102 is “1”, so that the process proceeds to step s360.

Next, at step s360, the RTIPSI
The G-multi module 605 generates a new ACCEPT message 125. First, the protocol number field 7
03 stores the protocol number of RTIPSIG-multi. In the destination address field 704, the value of the preceding hop node address field 809 of the routing table entry 802 retrieved as a result of step s300 is set. Further, a value indicating the ACCEPT message 502 is set in the command field 705. The next-hop node address field 709 is set to the value of the next-hop node address field 812 of the next-hop node entry 811 retrieved as a result of step s300. Stream ID field 707, QoS parameter field 706, transmission node address field 7
10, receiving node address field group 711, 712
Are the stream ID field 803, QoS parameter field 807, transmission node address field 804,
.., 80
Set according to the value of 5-n.

Next, in step s370, the RTIPSI
The G-multi module 605 activates the IP protocol stack module 604 to generate the packet generated in step s360, and passes the generated packet to the IP protocol stack module 604.

Next, referring to FIG.
The operation of receiving the ACCEPT message 125 from the router 102 will be described. Step s300 in FIG.
The processing of s310, s320, and s330 is performed by the router 102.
Is the same as the processing in.

Then, in step s340, the RTIPSIG-multi module 605 of the router 101 sends the previous hop node address field 80 of the routing table entry 802 retrieved as a result of step s300.
It is searched whether the value of 9 is “0”. Sending node 1
In step S350, since it is “0” in step S350
, The application is notified of the success of the resource reservation and the process is completed.

Next, referring to FIG. 10 to FIG. 16, when the application 606 on the transmission node 101 issues a simultaneous delivery request,
3 and 105 and the operations of the receiving nodes 104 and 106 will be described. First, the concept of the simultaneous delivery processing by the data delivery method of the present embodiment will be described with reference to FIG. The same reference numerals as those in FIG. 1 indicate the same parts.
FIG. 10 is a system block diagram illustrating simultaneous delivery processing using the data delivery method according to one embodiment of the present invention.

In this embodiment, the transmitting node 101
For the receiving node 104 and the receiving node 106,
IP compatible packets (continuous media data) are delivered simultaneously.

The transmitting node 101 transmits an IP compatible packet 130 to the router 102 which is an adjacent router of the transmitting node 101. The router 102 receives the IP compatible packet 14
01 is generated without performing memory copy (this method will be described later), and the IP compatible packets 131 and 132 are transferred to the router 103 and the router 105. The routers 103 and 105, which are adjacent routers of the receiving node, perform the conversion of the IP packet format from the IP compatible packets 131 and 132, and then transfer the IP packets 133 and 134 to the receiving node 104 and the receiving node 106. Thereby, the simultaneous delivery of the IP compatible packet is completed.

Next, the module configuration of the transmitting node 101 and the routers 102, 103, and 105 for executing the simultaneous delivery of the packet shown in FIG. 10 will be described with reference to FIG. Note that the same reference numerals as those in FIG. 2 indicate the same parts. FIG. 11 is a block diagram showing a module configuration of a transmission node and a router for simultaneous delivery in the data delivery method according to one embodiment of the present invention.

The present embodiment differs from the module configuration shown in FIG. 2 in that the module for receiving the packet received by the network receiving module 603 is not the IP protocol stack module 604 but the RTIP protocol stack module 1501. are doing.

RTIP protocol stack module 1
100 is an RTIP receiving module 1110, multi
An i relay module 1120 and a multi buffer release module 1130 are provided. When receiving the IP compatible packet, the network interface receiving module 603 activates the RTIP receiving module 1110 and passes the received packet. RTIP receiving module 1110
Selects only IP compatible packets for simultaneous delivery from received packets (this method will be described later).
Transfer to the relay / transmission module 1120. multi
The relay / transmission module 1120 is provided with the RTIPSI shown in FIG.
With reference to the RTIP routing table 610 constructed by the G-multi protocol stack module 605, an output-side network interface group 608 to output the corresponding packet is selected. In order to output a packet from the network interface group, the network interface transmission module group 607 is activated. When the packet transmission processing is completed by the network interface transmission module 607, the multi buffer release module 1130 is activated. The multi buffer release module 1130 executes a buffer release process (the details of this process will also be described later).

Application 6 on sending node 101
06 issues a packet simultaneous delivery request to the multi relay / transmission module 1120 using the following interface. <Function name> rt_multisend (stream_id, b
uf, buf_size) <argument> stream_id: Specifies the stream ID obtained by the return value of the allocate_resource function. buf: Specifies the start address of the user buffer. buf_size: Specifies the user buffer size. <Description> The user buffer data specified by buf and buf_size is simultaneously delivered to the receiving node group specified by the allocate_resource function according to the QoS parameter specified by the allocate_resource function.

The network interface transmission module 607 executes packet delivery strictly according to the QoS parameters described in the RTIP routing table 610. This shipping method is, for example, RTIP and RTIP
It is the same as the method using the SIG, and is described in "Osamu Takeuchi et al.," Design and Implementation of QoS Guaranteed Routing Method Applying Isochronous Scheduler ", Proceedings of Multimedia Communication and Distributed Processing Workshop, pp. 119-126." Use

Next, referring to FIGS. 12 and 13, FIG.
The data structure used by the transmission node 101 and the routers 102, 103, and 105 to realize the simultaneous delivery of the packet described in (1) will be described. First, the packet format of the IP compatible packet will be described with reference to FIG. FIG. 12 is an explanatory diagram of a packet format of an IP compatible packet for simultaneous delivery in the data delivery method according to one embodiment of the present invention.

The IP compatible packet has an IP header 160
1, a UDP header section 1604, and a data section 1605. The format of the IP compatible packet is the same as the existing UDP / IP packet.

In the IP header section 1601, the TOS field 1602 and the destination address field 1603
It is used in the packet simultaneous delivery processing shown in FIG. The TOS field 1602 stores the stream ID of the stream to which the corresponding IP compatible packet belongs.
The destination address field 1603 specifies whether or not the IP compatible packet is a packet for simultaneous delivery.
For simultaneous delivery packets, destination address field 1
603 includes an RTIP multicast address (22
4.0.0.0 to 239.255.255.255). In the case of a single receiving node delivery packet, other addresses are stored.

Next, the data structures of the command structure and the buffer structure will be described with reference to FIG.

The command structure 1701 is used to specify a packet to be transmitted when the multi relay / transmission module 1503 requests the network interface transmission module 607 to transmit a packet. The command structure 1701 contains the header field 17
02, a header size field 1703, and a buffer structure pointer field 1704.
The header field 1702 contains the header part 1 of the packet.
A pointer to a header area 1705 for storing 601 and 1604 is stored. The size of the header area 1705 is stored in the header size field 1703. The buffer structure pointer field 1704 stores a pointer to a buffer structure 1706 described later.

The buffer structure 1706 manages the use status of the buffer area 1710 for storing the data section 1605 of the packet. In accordance with this use status, the multi buffer release module 1504
It is determined whether to release 710.

The buffer structure 1706 includes a reference counter field 1707 and a buffer field 1708.
And a buffer size field 1709. The reference counter field 1707 stores the number of command structures 1701 that refer to the buffer structure (the buffer structure pointer field 1704 stores a pointer to the buffer). In the example of FIG. 13, since the buffer structure 1706 is referenced from the two command structures 1701, “2” is stored in the reference counter field 1707. The buffer field 1708 stores a pointer to the buffer area 1710. The size of the buffer area 1710 is stored in the buffer size field 1711.

Next, referring to FIG. 14, the R shown in FIG.
The operation of the TIP receiving module 1110 will be described. FIG. 14 is a flowchart showing the operation of the RTIP receiving module for simultaneous delivery in the data delivery method according to one embodiment of the present invention.

In step s400, the RTIP receiving module 1110 searches the destination address field 1603 of the received packet to determine whether or not the stored address is an RTIP multicast address. If it is not an RTIP multicast, a normal RTIP routing process is executed. If it is an RTIP multicast address, the process jumps to step s410.

Next, at step s410, the RTI
The P receiving module 1110 generates the buffer structure 1706 shown in FIG. After allocating a memory area to store the buffer structure, the reference counter field 1
707 is initialized to “0”. The buffer field 1708 and the buffer size field 1709 are also initialized so as to indicate the buffer passed from the network interface receiving module 603. Next, in step s420, the RTIP receiving module 1110 calls the multi relay / transmission module 1120 using the buffer structure 1706 generated in step s410 as an argument.

Next, referring to FIG. 15, the m shown in FIG.
The operation of the multi relay / transmission module 1120 will be described. FIG. 15 is a flowchart illustrating an operation of the multi-relay / transmission module for simultaneous delivery in the data delivery method according to an embodiment of the present invention.

In step s500, the multi-relay / transmission module 1120 checks the TOS of the received packet.
RTIP routing table entry 802 corresponding to the stream ID stored in field 1602
Is obtained using the stream ID index 801.

Next, the RTI obtained in step s500
Steps s510, s520, and s530 are executed for each next-hop node entry 811 connected to the P routing table entry 802. At first,
In step s510, the multi relay / transmission module 1120 secures a command structure 1701 and a header area 1705. In addition, header field 17
02, the header size field 1703 is initialized so as to indicate the secured header area 1705. Also, the buffer structure pointer field 1704 is initialized to point to the corresponding buffer structure 1707.
In addition, the value of the reference counter field 1708 of the indicated buffer structure 1707 is incremented.

Next, in step s520, mul
The ti relay / transmission module 1120 stores the header portions 1601 and 1604 of the received packet in the header area 1705.
Copy to Also, the I stored in the header area 1705
The value of the TOS field 1602 of the P header section 1601 is updated to the value of the stream ID field 813 of the focused next hop node entry 811. Further, if the local bit field 819 of the focused next hop node entry 812 is “1”, the value of the destination address field 1603 of the IP header section 1601 stored in the header area 1705 is changed to the corresponding next hop node entry. Next hop node address 812 of 811
Update to the value of Next, in step s530, the network interface transmission module 607 is activated using the command structure generated in step s520 as an argument.

The network interface transmission module 607 uses the DMA
Send the data in. Therefore, in the present embodiment, it is possible to transmit a packet to a plurality of network interfaces without copying the data body except for the header.

Next, referring to FIG. 16, the m shown in FIG.
The operation of the multi buffer release module 1130 will be described. FIG. 16 is a flowchart illustrating the operation of the multi buffer release module for simultaneous delivery in the data delivery method according to an embodiment of the present invention.

Multi buffer release module 113
0 is the network interface transmission module 60
7 is activated when packet transmission is completed. The network interface transmission module 607 is
The command structure passed by the ii relay / transmission module 1120 to the network interface transmission module 607 is also notified at the time of the activation.

First, in step s600, the header area 1705 specified from the command structure 1701 specified at the time of the notification is released.

Next, in step s610, the buffer structure 1706 pointed to by the command structure 1701
The value of the reference counter field 1707 is decremented. If the corresponding reference counter field 1707 is “0”, the corresponding buffer structure 1706 and buffer area 1708 are released. Next, step s62
At 0, the command structure 1701 is released.

As described above, according to the present embodiment, the transmitting node 101 only needs to transmit the IP compatible packet to the next hop node group which is smaller in number than the receiving node group. Even if the number of receiving nodes increases, the amount of resources to be reserved and the processing capacity consumed by the transmitting node 101 for simultaneous delivery processing hardly increase. Further, the ACCEPT message at the time of resource amount reservation is compiled for each router, and when all the ACCEPT messages are received, the message is transmitted to the preceding stage, so that the processing capacity consumed by the transmission node 101 at the time of reservation processing is reduced. It hardly increases. In addition, the router
Since packets are output to multiple network interfaces by DMA without performing memory copying, even if the number of networks to which continuous media data should be sent simultaneously increases, the router consumes data to send them to the corresponding network group. The processing capacity to be hardly increased. Further, simultaneous delivery of an IP compatible packet having the same format as that of an existing IP packet to a plurality of receiving nodes is realized.
Is guaranteed by the network interface transmitting module, so that simultaneous delivery of continuous media data can be executed to a plurality of receiving nodes having only the existing IP protocol stack.

[0104]

According to the present invention, even if the number of receiving nodes serving as simultaneous delivery destinations increases, the processing capacity consumed by the server for resource amount reservation processing and simultaneous delivery processing hardly increases. is there. Further, even if the number of networks to which continuous media data is to be transmitted increases at the same time, the processing capacity consumed by the router for transmitting data to the network group hardly increases.

[Brief description of the drawings]

FIG. 1 is a system block diagram showing a configuration of a data delivery system that realizes a data delivery method according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a module configuration of a transmission node and a router used in the data delivery method according to the embodiment.

FIG. 3 is an explanatory diagram of a data structure used by each module in the data delivery method according to the embodiment.

FIG. 4 is an RT used in the data delivery method according to the embodiment;
FIG. 4 is an explanatory diagram of a data structure of an IP routing table.

FIG. 5 is a flowchart illustrating an operation of transmitting a CONNECT message in the RTIPSIG-multi protocol stack module in the data delivery method according to an embodiment of the present invention.

FIG. 6 is a flowchart illustrating an operation of transmitting a CONNECT message in the RTIPSIG-multi protocol stack module in the data delivery method according to an embodiment of the present invention.

FIG. 7 is a flowchart illustrating an operation of transmitting and receiving a CONNECT message and transmitting an ACCEPT message in the RTIPSIG-multi protocol stack module in the data delivery method according to the embodiment of the present invention.

FIG. 8 is a flowchart illustrating an operation of receiving a CONNECT message in the RTIPSIG-multi protocol stack module in the data delivery method according to an embodiment of the present invention.

FIG. 9 is a flowchart illustrating an ACCEPT message receiving operation in the RTIPSIG-multi protocol stack module in the data delivery method according to one embodiment of the present invention.

FIG. 10 is a system block diagram illustrating simultaneous delivery processing using a data delivery method according to an embodiment of the present invention.

FIG. 11 is a block diagram showing a module configuration of a transmission node and a router for simultaneous delivery in a data delivery method according to an embodiment of the present invention.

FIG. 12 is an explanatory diagram of a packet format of an IP compatible packet for simultaneous delivery in a data delivery method according to an embodiment of the present invention.

FIG. 13 is an explanatory diagram of a data structure of a command structure and a buffer structure for simultaneous delivery in a data delivery method according to an embodiment of the present invention.

FIG. 14 is a flowchart showing the operation of the RTIP receiving module for simultaneous delivery in the data delivery method according to one embodiment of the present invention.

FIG. 15 is a flowchart illustrating an operation of a multi-relay / transmission module for simultaneous delivery in a data delivery method according to an embodiment of the present invention.

FIG. 16 is a flowchart illustrating an operation of a multi-buffer release module for simultaneous delivery in a data delivery method according to an embodiment of the present invention.

[Explanation of symbols]

 101 transmitting node 102, 103, 105 router 104, 106 receiving node 605 RTIPSIG-multi protocol stack module 610 RTIP routing table 1100 RTIP protocol stack module 1110 RTIP receiving module 1120 multi relay / receiving module 1130 multi buffer release module

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Shoji Kodama 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture F-term in Hitachi, Ltd. System Development Laboratory Co., Ltd. F-term (reference) 5K030 GA08 HC14 HD03 JA11 LD02 5K033 CB13 DB14 DB18 EC03

Claims (3)

[Claims] 1. A data delivery method for performing resource reservation processing from an information transmitting apparatus to a plurality of information receiving apparatuses via a plurality of information relay apparatuses and then simultaneously delivering data. Identification information for identifying the plurality of information receiving apparatuses requesting data reception between the information transmitting apparatus and the information relay apparatus belonging to the same network as the plurality of information receiving apparatuses, and requesting the information transmitting apparatus Transmitting and receiving control information relating to a data transfer rate; B) the information transmitting device and the information relay device receiving the control information reserving resources required for data transfer; and C) the information. Identification information for identifying the information relay device or the information receiving device that the transmitting device and the information relay device that has received the control information should be the next relay destination of data Constructing; D) upon completion of the resource reservation, after receiving a resource reservation end message from all of the information relay devices in the next stage, and in the preceding stage, the information relay device or the information transmitting device Transmitting a resource reservation end message, and performing resource reservation processing. 2. A data delivery method for performing resource reservation processing from an information transmitting apparatus to a plurality of information receiving apparatuses via a plurality of information relay apparatuses and then simultaneously delivering data. The information transmitting device reads the identification information for identifying the information relay device or the information receiving device built in the information relay device by the resource reservation processing, and requests the read information relay device or information receiving device to transmit the requested data to the information relay device or the information receiving device. F) transmitting the data according to the transfer rate; and F) the information relay device receiving the transmitted data reads the identification information for identifying the information relay device or the information receiving device to be the next relay destination, and Sends to the network interface group that connects the information relay device or information receiving device to be the next relay destination read out Data and a step of requesting to transmit using DMA, and data delivery method and executes the simultaneous delivery processing of the data. 3. The data delivery method according to claim 2, wherein: G) the information relay device belonging to the same network as the information receiving device that has received the transmitted data has a data format in which the transmitted data is stored. Converting the data into a data format to be used when data is transmitted only from the information transmitting device to the information receiving device.