JP2004221756A - Information processing apparatus and information processing method, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an apparatus and method with a distribution type data storage processing configuration for efficiently executing processing of acquiring and reproducing data from each node. <P>SOLUTION: A reproduction instruction apparatus for transmitting a data reproduction processing request to a node connected to a network is configured to revise and set a reply probability β to be applied to the decision processing for deciding whether or not the node receiving the reproduction processing request packet executes the processing in compliance with the processing request depending on a packet transmission sequence and to control the transmission timing of the reproduction processing request packet. The apparatus sets a higher reply probability β for an initial data request, executes the processing of setting a lower reply probability β to prevent reception of wasteful packets when a buffer free capacity is decreased and also adjusts the packet transmission interval. Thereby, the efficient utilization of the buffer and high speed reception are realized. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an information processing apparatus, an information processing method, and a computer program. More specifically, in a distributed storage processing configuration in which data is distributedly recorded in a plurality of nodes, an information processing apparatus and information enabling efficient data acquisition and reproduction processing when executing data acquisition / reproduction processing from each node The present invention relates to a processing method and a computer program.
[0002]
[Prior art]
In recent years, various information processing apparatuses, for example, personal computers (PCs), large computers, servers, and other communication devices have been connected to a communication network such as the Internet, and video, image data, audio data, Alternatively, transfer of contents such as various programs or transfer of various processing data is performed. The type of content exchanged through a network is shifting from text and still images to multimedia contents such as moving images and audio, and the capacity of the content is remarkably increasing.
[0003]
Attention has been paid to a large-scale storage system for distributing and recording certain data in a large number of information processing terminals connected to each other via a network. In such a distributed storage system, a server that records and manages data transmits data to an information processing terminal or another server by multicast or the like, and stores the data in a local storage provided in the information processing terminal or another server. Recorded on medium.
[0004]
In this case, in order to be able to retrieve data on demand, a large amount of data must be recorded on a recording medium. For example, in the case of a movie having a data capacity of about 2 gigabytes per movie, if 500 such video data are recorded, a capacity of 1 terabyte or more is required.
[0005]
In addition, when the server provides data to the client requesting the data in a unicast manner by streaming, in order to perform error-free transmission, for example, a TCP / IP arriving signal (ACK) is used. A protocol that requires retransmission of data is used.
[0006]
However, this method imposes a heavy burden on the server side, so that even if one high-performance server is used, at present, services can be provided only to several hundred clients. Even if a protocol that does not use ACK such as UDP / IP is used, the number of clients that can be serviced is about several thousands. As described above, when data is provided by streaming, the cost of the server increases, and the number of clients is limited.
[0007]
In a conventional streaming system, a system for streaming a stream from a single server or a cache server is generally used. In such a system, the server side transmits the data to the network at the same timing, and the client side executes the PLL control to control the reproduction timing. In order to execute the PLL control, it is necessary to provide a buffer for absorbing the jitter of the arrival time. The input data is temporarily stored in a buffer on the client side that receives data, and the data is extracted from the buffer. Therefore, it is necessary to perform reproduction by PLL control. Therefore, there has been a delay from the reception of the image or the sound to the actual output. In addition, if the buffer capacity of the client is not sufficient, there is a problem that the buffer overflows and a reproduction error occurs.
[0008]
In recent years, Microsoft's WMP9 (Corona) and Real Media's True Stream have adopted a method based on feedback from a client to a server as a method for improving the buffer capacity problem on the client side. However, when such a feedback method is executed, when data is distributed to a large number of clients, requests from a large number of clients to the server are generated, and a load is imposed on the server. Therefore, in a system in which there are a large number, for example, thousands or tens of thousands of playback clients such as data distribution via the Internet, unless an infrastructure is constructed such as arranging a large number of servers, cache servers, etc. on a network, The solution by the feedback method is difficult.
[0009]
Therefore, a method of distributing the load on the server has been proposed. For example, there is a method in which a node in which a cache exists is specified in advance to avoid load concentration on a specific server. However, in this approach, it is necessary to place a search server in the center, execute an inquiry to the search server, and then access the content server in order to identify the node where the cache exists in advance. Delay occurs.
[0010]
Therefore, in recent years, a method of transmitting data to a plurality of clients without requesting retransmission of data by using FEC (Forward Error Correction) as a multicast technique has been proposed. In this method, a server repeatedly transmits a stream by multicast, and a client picks up a necessary signal from the stream, decodes the picked-up data, and reproduces the data.
[0011]
In order to transmit 500 video data of 2 Gigabytes per movie within 10 minutes using this method, a transmission band of about 14.7 Gbit / s is required. Furthermore, when transmitting the same amount of video data within one minute, a transmission band of about 147 Gbit / s is required. This is a theoretical value, but a server that can withstand such a capacity and transmission method is very costly and impractical even if realized. In addition, there is a method of distributing and recording data on multiple hosts, but in order to realize this system, huge data must be managed by multiple servers, so data management and data communication are required. Processing increases.
[0012]
In recent years, peer-to-peer (P2P: Peer-to-Peer) network technology has been developed and used as a direct communication process between information processing apparatuses. The P2P network does not include a server that performs intensive processing, but includes an information processing device as a resource of each network client, such as a PC, a mobile terminal, a PDA, a mobile phone, and a function capable of performing communication processing. Various devices, such as a disk device or a printer, as storage means connected to or connected to a communication device, communicate with each other via a network, and the resources of each network client can be shared.
[0013]
Peer-to-Peer (P2P: Peer-to-Peer) network technology is said to have been used for the first time in Advanced Peer-to-Peer Networking (APPN) proposed by IBM Corporation. By using this network, there is no need to install a huge distribution server required for distributing contents in a conventional client-server type network, and contents distributed to resources of each network client can be distributed. Many users are available, and large-capacity content can be distributed and stored.
[0014]
Peer-to-peer (P2P) networks include "Pure-Peer-to-Peer (P2P) networks" and "Hybrid (peer-to-peer) networks". (P2P: Peer-to-Peer network).
[0015]
A pure peer-to-peer (P2P) network is a network form in which each component (peer) of the system has an equal function / role and performs equal communication. A typical service using this is Gnutella. A hybrid peer-to-peer (P2P) network is a pure peer-to-peer (P2P) network as well as each component of the system. (Peer: Peer) It is a network form that uses a control server to facilitate interaction. A typical service using the service is Napster.
[0016]
In a hybrid peer-to-peer (Hybrid P2P) system represented by Napster, when a network-connected terminal attempts to acquire content, first, a content server is searched for a content resource by a central server. Then, based on the search information, access is made to the node (other network connection terminal) holding the resource to obtain the content. In this method, it is necessary to register the resource information of all nodes in the central server, and furthermore, there is a disadvantage that the search is concentrated on the central server.
[0017]
Therefore, a method has been proposed in which processes such as resource search are distributed to a plurality of devices and executed. In this processing distribution method, processing determination apparatuses are managed by a method such as arranging them in a tree-like relationship, and processing such as resource search is distributed to a plurality of apparatuses and executed based on management information. However, even in this method, when the number of devices that execute processing increases, for example, to several millions, the amount of information for managing the tree configuration increases, the number of processing instructions for transmitting execution instructions to a plurality of processing devices increases, or the tree increases. Problems such as ensuring the consistency of the system arise. In addition, there is a problem that a processing delay occurs because determination processing by a plurality of processing execution determination devices is required.
[0018]
In order to compensate for these weaknesses, there is a method in which all commands are sent to all network connection nodes, and each node determines whether or not to execute the received processing command. This is a pure peer-to-peer (Pure P2P) system represented by Gnutella. This scheme is different from the hybrid peer-to-peer (Hybrid P2P) scheme in that it does not have a central server that executes a resource search process. Is executed to request the hit terminal for a processing request such as content transmission.
[0019]
Also in the pure peer-to-peer (Pure P2P) system represented by Gnutella, all or as much as possible of the search instruction is transferred by using a routing such as a tree structure or a network structure. A configuration in which a node performs a search is effective. However, this method also has a drawback in that an instruction transfer process of a processing instruction that is not executed in its own node is executed, and a load is imposed on a transmission path.
[0020]
For example, in order to search all the network connection nodes and make the processing request reach all the nodes, complicated routing management is required. On the other hand, when the node search of the best effort method is executed, it is not guaranteed that the instruction is transmitted to all the nodes, and a necessary resource may not be found. Further, there is a problem that congestion in the network occurs when communication for node search occurs frequently.
[0021]
[Problems to be solved by the invention]
The present invention has been made in view of the above-described various problems in the related art, and has a network configuration in which a large number of terminals (nodes) are connected, that is, a distributed storage process in which data is distributed to a plurality of nodes and recorded. It is an object of the present invention to provide an information processing apparatus, an information processing method, and a computer program which can efficiently acquire distributed data and perform a reliable reproduction process based on the acquired data.
[0022]
More specifically, an information processing apparatus and an information processing apparatus that increase the efficiency of use of a buffer of a reproduction request node as a reproduction instruction device, reduce the time loss from data reception to reproduction processing, and enable efficient data reproduction. A processing method and a computer program are provided.
[0023]
According to the present invention, when performing data reproduction by acquiring data stored in storage means installed in a plurality of terminals (nodes) connected to a network, a reproduction request can be made without performing a process such as a node search in advance. Is transmitted to the terminal (node) that has received the request. Under the distributed data processing configuration in which it is determined whether or not to execute the instruction autonomously and the instruction is selectively executed, the reproduction request node has It is an object of the present invention to provide an information processing apparatus, an information processing method, and a computer program that can execute efficient data acquisition and reproduction processing by executing feedback control such as a retransmission request according to the capacity of a buffer.
[0024]
[Means for Solving the Problems]
According to a first aspect of the present invention,
An information processing device as a reproduction instruction device for transmitting a data reproduction processing request to a node connected to a network,
A packet generation unit that generates a data reproduction processing request packet;
A rule for changing and setting the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence A judgment condition setting unit;
A phase control unit that controls a transmission timing of the data reproduction processing request packet, a network interface unit that transmits the packet generated by the packet generation unit,
And a buffer for storing packet storage data received from the node as a response to the data reproduction processing request packet.
[0025]
Further, in one embodiment of the information processing apparatus of the present invention, the rule determination condition setting unit sets the value of the return probability β as a gradually reduced value in a data reproduction process request packet transmission sequence to be continuously transmitted. It is characterized by having a configuration for executing a process to perform.
[0026]
Further, in one embodiment of the information processing apparatus of the present invention, the phase control unit executes control processing for setting a data reproduction processing request packet transmission interval to be gradually longer in a data reproduction processing request packet transmission sequence to be continuously transmitted. It is characterized by having the structure which does.
[0027]
Further, in one embodiment of the information processing apparatus of the present invention, the data to be reproduced stored in the node is obtained by dividing the data into p blocks and subjecting the p blocks to FEC encoding to obtain q blocks. The rule determination condition setting unit is coding processing data of the coding rate q / p converted into blocks, and the rule determining condition setting unit determines a probability value for returning data at the node receiving the data reproduction processing request with a return probability β. This is a configuration that is set as a condition description. The recording probability α specified by the recording instruction device connected to the network, the number of coding blocks q, and the number of return blocks q × α that can be calculated from the number n of network connection nodes The relationship between × n × β and the number of blocks p is the number of reply blocks: q × α × n × β> the number of blocks: p. When the remaining capacity is large, the value of the reply probability β set in the data reproduction processing request packet is set to be high, and when the remaining capacity of the buffer is small, control is performed to set the value of the reply probability β to be low. Features.
[0028]
Further, in one embodiment of the information processing apparatus of the present invention, the phase control unit determines that the reception response packet corresponding to the data reproduction processing request as the previous request has reached the value p which is the number of packets that can be decoded. A phase control for transmitting the next request is performed as a condition or on condition that the number becomes close to the value p.
[0029]
Further, in one embodiment of the information processing apparatus of the present invention, the information processing apparatus further includes a data restoration processing unit that performs a deinterleaving process and an FEC decoding process, and the data restoration processing unit includes the data reproduction unit. The deinterleaving process and the FEC decoding process are performed on data to be reproduced extracted from the packet received from the node that has received the processing request, and the data is restored.
[0030]
Further, in one embodiment of the information processing apparatus of the present invention, the information processing apparatus further includes a timing synchronization processing unit that receives synchronization control information from the outside and controls transmission timing of the data reproduction processing request packet. It is characterized by having a configuration.
[0031]
Further, a second aspect of the present invention provides
An information processing method for performing data reproduction processing control based on transmission of a data reproduction processing request to a node connected to a network,
Rule determination for changing the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence Condition setting step;
A packet generation step of generating a data reproduction processing request packet;
A phase control step of controlling a transmission timing of the data reproduction processing request packet,
An information processing method comprising:
[0032]
Further, in one embodiment of the information processing method of the present invention, the rule determination condition setting step sets the value of the return probability β as a value that is gradually reduced in a data reproduction processing request packet transmission sequence that is transmitted continuously. Is performed.
[0033]
Further, in one embodiment of the information processing method of the present invention, the phase control step executes a control process of setting a data reproduction process request packet transmission interval to be gradually longer in a data reproduction process request packet transmission sequence to be continuously transmitted. It is characterized by doing.
[0034]
Further, in one embodiment of the information processing method according to the present invention, the data to be reproduced stored in the node is obtained by dividing the data into p blocks and subjecting the p blocks to FEC encoding to obtain q blocks. The coding process data of the coding rate q / p converted into blocks, wherein the rule determination condition setting step determines a probability value at which data is returned at the node receiving the data playback processing request with a return probability β. Set as a condition description, the recording probability α specified by the recording instruction device connected to the network, the number of coding blocks q, and the number of reply blocks q × α × n × The relationship between β and the number of blocks p is the number of reply blocks: q × α × n × β> the number of blocks: p. Be greater, the higher set value of the return probability beta to be set to the data reproduction process request packet, if the buffer remaining capacity is small and performs a control to set a low value of the return probability beta.
[0035]
Further, in one embodiment of the information processing method according to the present invention, the phase control step includes a step of determining that a reception response packet to the data reproduction processing request as the previous request has reached the value p which is the number of packets that can be decoded. The phase control for transmitting the next request is performed as a condition or on condition that the number becomes close to the value p.
[0036]
Further, in one embodiment of the information processing method of the present invention, the information processing method further includes a data restoration processing step of executing a deinterleave processing and an FEC decoding processing, and the data restoration processing step includes the data reproduction processing step. The present invention is characterized in that deinterleave processing and FEC decoding processing are performed on reproduction target data extracted from a packet received from a node that has received a processing request, and data restoration is performed.
[0037]
Further, in one embodiment of the information processing method of the present invention, the information processing method further includes a timing synchronization processing step of receiving synchronization control information from the outside and controlling transmission timing of the data reproduction processing request packet. It is characterized by the following.
[0038]
Further, a third aspect of the present invention provides
A computer program that executes data reproduction control processing based on transmission of a data reproduction processing request to a node connected to a network,
Rule determination for changing the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence A condition setting step;
A packet generation step of generating a data reproduction processing request packet;
A phase control step of controlling a transmission timing of the data reproduction processing request packet,
A computer program characterized by having:
[0039]
[Action]
According to the configuration of the present invention, in the reproduction instruction device that transmits a data reproduction processing request to a node connected to a network, it is determined whether or not the receiving node of the reproduction processing request packet executes the processing according to the processing request. The return probability β to be applied to the processing is changed and set according to the packet transmission sequence, and the transmission timing of the data reproduction processing request packet is controlled. At first, it was allowed to increase the useless packets to fill the buffer quickly. When a high return probability β value is set and the buffer free space decreases, a process of setting a low return probability β is performed to prevent the reception of useless packets, and by adjusting the packet transmission interval, It is possible to perform data reproduction that realizes efficient use of the buffer and high-speed reception.
[0040]
Further, according to the configuration of the present invention, the recording probability α specified by the recording instruction device connected to the network, the number of coded blocks q, and the number of return blocks q × α that can be calculated from the number of network connection nodes n When the relationship between × n × β and the number of blocks p is the number of reply blocks: q × α × n × β> the number of blocks: p, the return probability: β is used as a reference value, and when the remaining buffer capacity is large, The value of the reply probability β set in the data reproduction processing request packet is set to a high value, and if the remaining buffer capacity is small, the control is performed to set the value of the reply probability β to a low value. As a result, the buffer use efficiency is improved, and high-speed reception is realized.
[0041]
Further, according to the configuration of the present invention, it becomes possible for the reproduction instruction device to control the transmission timing of the reproduction command based on the timing control packet received from the timing generation device, and execute the recording process and the reproduction process in parallel. In this case, the reproduction process can be performed in accordance with the progress of the recording process without causing a recording delay or a reproduction process delay.
[0042]
The computer program of the present invention is provided, for example, in a computer-readable format for a general-purpose computer system capable of executing various program codes, in a storage medium or communication medium such as a CD, FD, or MO. Or a computer program that can be provided by a communication medium such as a network. By providing such a program in a computer-readable format, processing according to the program is realized on a computer system.
[0043]
Further objects, features and advantages of the present invention will become apparent from the following detailed description based on embodiments of the present invention and the accompanying drawings. In this specification, the term “system” refers to a logical set of a plurality of devices, and is not limited to a device having each component in the same housing.
[0044]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an information processing apparatus, an information processing method, and a computer program according to the present invention will be described in detail with reference to the drawings. The description will be made in the following order.
1. Network configuration
2. Information processing device configuration as a node
3. Recording instruction device configuration and data recording process
4. Playback instruction device configuration and data playback processing
5. Data recording / playback processing sequence
6. Synchronous processing during recording and playback
7. Hardware configuration of information processing device
[0045]
[1. Network configuration]
First, an example of a network configuration that realizes data processing to which the information processing apparatus of the present invention is applied will be described with reference to FIG. FIG. 1 is a diagram showing a specific example of a distributed storage system for distributing and recording data to terminals constituting a network. The distributed storage system includes n nodes 111, 112,..., And 115 connected to a communication network of the network 100, a recording instruction device 101 that instructs and controls recording of data in each node, and a recording instruction device 101 that is recorded in each node. And a reproduction instructing device 102 for reading the read data.
[0046]
In this embodiment, the recording instruction apparatus 101, the nodes 111 to 115, and the reproduction instruction apparatus 102 will be described as separate apparatuses. However, an apparatus such as a recording / reproduction instruction apparatus having the functions of these two apparatuses will be described. There may be. Further, each of the nodes may be provided with a function as a recording instruction device or a reproduction instruction device. In this case, each device constituting the network can be used without distinction. Note that the recording instruction device, the reproduction instruction device, and the node are collectively called an information processing device.
[0047]
That is, a network configuration as shown in FIG. 2 can be assumed. The information processing device has a configuration capable of communicating with another information processing device and is connected by a communication network. The network is basically a pure peer-to-peer (P2P) network that does not have a control server as a look-up server (look-up server).
[0048]
As shown in FIG. 2, the information processing devices 121 to 126 mutually communicate via the network 110, and execute various data processes such as a content recording process and a content reproduction process. The present invention is also applicable to a hybrid peer-to-peer (P2P) network in which a control server exists, and is not particularly applicable to a limited network configuration. Instead, any configuration is possible as long as the information processing apparatuses can communicate with each other.
[0049]
One of the information processing devices 121 to 126 shown in FIG. 2 outputs a content recording request or a content reproduction request to an information processing device connected to the network. When a content recording request is made, for example, a recording request processing request packet in which a multicast address is set based on a predetermined rule is generated, and the packet is subjected to FEC (Forward Error Correction) and stored as interleaved divided data. , To another information processing apparatus. The information processing apparatus that receives these packets autonomously determines whether or not to execute data recording processing, and performs content recording processing based on the determination.
[0050]
When the content is to be reproduced, one of the information processing apparatuses 121 to 126 generates a content reproduction request and transmits the generated packet to another information processing apparatus, for example, a request packet in which a multicast address based on a predetermined rule is set. . The information processing apparatus that receives these packets autonomously determines whether or not to execute data extraction processing, performs content extraction processing based on the determination, and extracts data, for example, FEC (Forward Error Correction) according to the request. And the interleaved divided data is stored in a packet and transmitted to the reproduction request information processing apparatus.
[0051]
Although a transmission control device such as a router constituting a network is omitted in FIGS. 1 and 2, a transmission control device such as a router for selecting a route of a packet passing through a node is actually provided. ing. The transmission control device may be provided separately from the node, or the node may have a function as the transmission control device.
[0052]
[2. Information processing device configuration as a node]
Next, the configuration of the information processing device functioning as a node will be described. FIG. 3 is a diagram illustrating a configuration of an information processing apparatus that functions as a node.
[0053]
As shown in FIG. 3, the information processing device (node) 200 includes a rule determination processing unit 201, a data processing unit 202, a packet processing unit 203, and a data transmission / reception unit 204. The data transmission / reception unit 204 outputs a transmission packet to another information processing device connected via a network and executes a process of inputting a reception packet from the other information processing device.
[0054]
The packet processing unit 203 performs a process of generating a packet to be transmitted from the own device to another information processing device connected to the network, a process of analyzing a packet received from another information processing device, and the like.
[0055]
The data processing unit 202 executes a process according to a data processing program stored in each information processing device. For example, data processing based on a data processing request from an information processing device such as another recording instruction device or a reproduction instruction device connected to a network, for example, in the case of a content reproduction request, the specified content is taken out from the storage unit 205, This is processing to be output to the processing unit 203. In the case of a content recording request, the process is to input and store the input content in the storage unit 205.
[0056]
The rule determination processing unit 201 is a processing unit that executes processing unique to the information processing apparatus according to the present invention, and based on a probability value: α or β as [rule determination condition description] described in an input packet. In addition, data processing based on a packet, for example, data reception and storage processing to a storage unit as processing corresponding to a data recording instruction, or data extraction and transmission processing from the storage unit as processing corresponding to a data reproduction instruction, these processings Is performed to determine whether or not to execute. In the configuration of the present invention, the probability of executing a recording command is set as α, and the probability of executing a playback command is set as β, and these probabilities are set in a recording request device or a recording request or a reproduction request packet transmitted by the recording device. The value is set.
[0057]
For example, when executing the content recording process, if the number n of the nodes configuring the distributed storage system is sufficiently large and the number q of the decoded blocks is sufficiently large, the packets are uniformly recorded in all the nodes. , Data is recorded with a probability of α for the entire distributed storage system.
[0058]
When executing the content reproduction processing, the rule determination processing unit 201 of each node determines whether or not to execute the reproduction instruction based on the probability: β and executes it. As described above, in the reproduction command, since the command is executed only at a certain probability in each node, packet loss occurs. However, if the probability β of the execution of the reproduction command is set so that a sufficient number of packets are sent from a plurality of distributed nodes in total, these are combined and based on FEC (Forward Error Collection). You can perform error correction and play back the original data.
[0059]
[3. Recording Instruction Device Configuration and Data Recording Process]
Next, the recording instruction device and the data recording process will be described. FIG. 4 is a diagram showing a configuration of the recording instruction device 250. The recording instruction device 250 executes data processing such as a data input unit 251 that inputs recording processing request data (contents) to a plurality of nodes, for example, FEC (Forward Error Collection) encoding, and interleave processing on encoded data. A data processing unit 252, a rule determination condition setting unit 253 for setting a probability value: α as the above-mentioned “rule determination condition description”, processed data generated by the data processing unit 252, and a rule determination set by the rule determination condition setting unit 253 It has a packet generator 254 for storing a condition description and generating a packet with an address set, and a network interface 255 for connecting to a network.
[0060]
Here, the FEC coding is a general term for coding methods that perform error correction on the receiving side, such as a tornado coding method, a read tornado coding method, and a turbo coding method, and the data processing unit 252 includes a data input unit 251. Is divided into p blocks, and the p blocks are subjected to FEC encoding to be converted into q blocks. Encoding from the p blocks to the q blocks is referred to as encoding of a coding rate q / p. By changing the coding rate q / p, the recording efficiency of the distributed storage system can be improved. The transmission efficiency can be changed.
[0061]
The interleaving process is a process of rearranging the order of encoded data. By interleaving, data can be dispersed, and a burst error caused by packet loss can be made a random error. As a result, a lost data portion can be corrected by error correction according to FEC. .
[0062]
The FEC encoding process and the interleaving process will be described with reference to FIG. As shown in FIG. 5A, the input original data is divided into p blocks. Then, as shown in FIG. 5B, FEC encoding is performed at an encoding rate of q / p, and the data divided into p blocks is converted into q encoded blocks.
[0063]
As described above, FEC coding is a general term for coding methods that perform error correction on the receiving side, such as a tornado coding method, a lead tornado coding method, and a turbo coding method. When certain data is coded at the coding rate q / p, as described in the paper RIZZ097 (http://www.et.unipi.it/@luigi/fec.html#fec.ps), If more than one coded block remains, the original message can be restored even if some blocks are lost.
[0064]
The data subjected to the FEC encoding is subjected to an interleaving process as shown in FIG. The order of the encoded data is rearranged to distribute the data. The data subjected to the interleave processing is output to the rule determination condition setting unit 253, and the probability value: α is set as the above [rule determination condition description].
[0065]
The setting process of the probability value: α as the [rule determination condition description] is set in relation to the above-described process of the data processing unit. The data processing unit executes coding processing of a coding rate q / p in which data is divided into p blocks, and the generated p blocks are subjected to FEC coding and converted into q blocks. In this case, the rule determination condition setting unit 253 is configured to set, as a recording rule determination condition description, a probability value: α at which data is to be recorded with a recording probability: α at a node that receives a data recording processing request, and is connected to a network. The return probability designated by the reproduced playback instruction device 102 (see FIG. 1): β, the number of coded blocks: q, and the number of reply blocks that can be calculated from the number of network connection nodes: n: q × α × n × The relationship between β and the number of blocks: p is
Number of reply blocks: q × α × n × β> Number of blocks: p
A probability value: α is set so that With this setting, reliable data restoration from return data returned from each node with a return probability: β in response to a reproduction request from the reproduction instruction device is guaranteed.
[0066]
The packet generation unit 254 divides the processed data generated by the data processing unit 252 into a predetermined size, stores the rule determination condition description set by the rule determination condition setting unit 253, and adds a header and a footer including an address. A packet is generated and transmitted via the network interface 255. The packet is transmitted to each node configuring the distributed storage system using the address setting according to the transmission node, that is, using unicast or multicast.
[0067]
FIG. 6 is a diagram illustrating a structure of the packet 40 generated by the packet generation unit 254. The packet is composed of a header, a description of a recording rule determination condition, a payload, and a footer. The payload stores processed data (processed content) on which FEC encoding processing and interleaving processing have been performed. In the header and footer, control information such as a data ID indicating a data type, a CRC (Cyclic Redundancy Check) checksum, a GUID (Global Unique ID) as a unique identifier of the packet transmission destination node, and a network address are described. I have.
[0068]
In the recording rule determination condition description, a probability that each node described below records this packet: α is described. The rule determination processing unit 201 (see FIG. 3) of the node that has received the packet records the packet based on the recording probability α.
[0069]
All nodes constituting the distributed storage system determine whether to record a packet based on the recording probability. As a result, data is recorded with the probability α in the nodes constituting the distributed storage system. In this distributed storage system, when the number n of nodes is sufficiently large and the number q of decoded blocks is sufficiently large, data can be distributed to each node with equal probability.
[0070]
The recording rule determination condition description is stored in a packet different from the packet storing the processed content, the link information of the packet storing the recording rule determination condition description is stored in the packet storing the processed content, and the processed content is stored. The node that has received the packet may acquire the corresponding recording rule determination condition description storage packet based on the link information.
[0071]
Next, a recording processing procedure in which a node that has received a packet autonomously determines whether or not to record content based on transmission of a recording processing request packet from the recording instruction device and executes the processing will be described.
[0072]
FIG. 7 is a diagram illustrating a process of transmitting the data recording processing request packet shown in FIG. 6 as a data recording command from the recording instruction device 101 to the plurality of nodes 111 to 115. The recording instruction device uses the processed data subjected to the FEC process and the interleave process described in FIG. 6 as a payload, and further transmits a packet in which a probability value: α as a rule determination condition description is set to each node by unicast or multicast. Send.
[0073]
With reference to FIG. 8, a processing procedure in the node that has received the packet will be described. First, in step S101, a data recording request packet is awaited. If it is determined in step S102 that the packet has been received, a random number generation process is executed in step S103. In step S104, the generated random number and the data recording process request Recording rule judgment condition description stored in the packet: comparison with the probability α is performed, and execution or non-execution of the instruction is judged based on the comparison result.
[0074]
For example, if the generated random number> probability α, the processing for recording the processed data stored as the payload of the packet in its own storage means is executed. If the generated random number ≦ probability α, the data recording processing is not executed.
[0075]
Note that, when the number n of the nodes constituting the distributed storage system is sufficiently large and the number q of the decoded blocks is sufficiently large, the random numbers generated in each node are such that the packets are equally transmitted to all the nodes. The range of the generated random number is set so that the data is recorded and the data is recorded with the probability of α as the entire distributed storage system.
[0076]
If it is determined that the data recording processing request should be executed as a result of the comparison processing based on the random number in step S104, the process proceeds to step S105, where processing for extracting the processed data stored as the payload of the packet is executed, and A process for recording the extracted data in the storage unit of the device (node) is executed. On the other hand, if it is determined that the data recording processing request is not to be executed as the comparison processing result based on the random number, the process proceeds to step S106, and the processing ends without executing the data recording processing.
[0077]
As described above, the information processing apparatus that has received the data recording processing request (instruction) autonomously determines the execution or non-execution of the instruction based on the comparison between the recording rule determination condition description stored in the packet and the random number. , And performs processing according to the determination.
[0078]
Next, referring to the processing flow of FIG. 9, the information processing apparatus that has received the data recording processing request calculates a hash value, and autonomously executes and does not execute the data recording processing request based on the hash value. Will be described.
[0079]
First, in step S201, a data processing request packet is awaited. If it is determined in step S202 that the packet has been received, a hash value generation process based on data stored in the received packet is executed in step S203. The execution or non-execution of the instruction is determined based on the hash value thus obtained.
[0080]
For example, the information processing apparatus stores a preset value as a preset threshold value in the storage unit, and executes the command if the generated hash value is greater than the set value, and executes the instruction if the generated hash value is smaller than or equal to the set value. Make settings such as execution.
[0081]
As shown in FIG. 10, the data stored in the data processing request packet for which the hash value is to be generated is, for example, an identifier (data ID) of the content data or a part of the data, for example, a predetermined (n) from the head of the data. Bit data can be set as hash target data. For example, MD5 can be applied to the hash value calculation, and a hash value based on MD5 of the data ID or a hash value based on MD5 of the data content is generated as shown in FIG.
[0082]
If the result of the determination based on the hash value in step S204 indicates that the data recording processing request should be executed, the process proceeds to step S205, where the processing for extracting the processed data stored as the payload of the packet is executed, and A process for recording the extracted data in the storage unit of the device (node) is executed. On the other hand, if it is determined that the data recording processing request is not to be executed as the comparison processing result based on the hash value, the process proceeds to step S206, and the processing ends without executing the data recording processing.
[0083]
As described above, the information processing apparatus that has received the data recording processing request (instruction) autonomously executes and non-executes the instruction based on the comparison between the hash value of the data stored in the packet and the set value of each node. Is determined, and processing according to the determination is performed.
[0084]
[4. Playback instruction device configuration and data playback processing]
Next, a reproduction instruction device 102 (see FIG. 1) that transmits a data reproduction instruction to each node, receives data to be reproduced from each node, and executes data reproduction processing will be described. FIG. 11 is a diagram showing the configuration of the reproduction instruction device 270. The reproduction instruction device 270 stores a network interface 271 for transmitting and receiving data to and from the outside via a network, a packet processing unit 272 for performing processing such as taking out a payload from an input packet, and data to be reproduced included in the payload. A buffer 273, a data restoration processing unit 274 that executes deinterleaving processing and FEC decoding processing of data extracted from a received packet and executes restoration of content data, and supplies the restored data to an external device such as a monitor or a speaker (not shown). A data processing unit 275 for executing output processing control, a rule determination condition setting unit 276 for determining a probability value: β as a [rule determination condition description] to be set in a data request packet, and playback data transmission in which a playback target data is specified for a node Generate a packet requesting Over data request packet generating unit 277, a phase control section (packet output timing control unit) 278 for executing phase control of the output timing control of the data request packet.
[0085]
The data request packet generator 277 transmits a packet requesting data to each node constituting the distributed storage system. FIG. 12 is a diagram showing a configuration of a packet requesting data. The packet includes a header, a playback rule determination condition description section, a request description section, and a footer. In the request description part, a data ID for identifying requested data is recorded. Control information such as a CRC checksum, a network address or GUID of a node, and a sequence number indicating the order of data is recorded in the header and footer.
[0086]
The return probability β determined by the rule determination condition setting unit 276 is set in the reproduction rule determination condition description unit. The return probability β is a variable for determining whether the node that has received the packet returns data. Based on this variable, some nodes determine that data is to be returned, and some nodes determine not to return data. The return probability β is a value when the entire distributed storage system is viewed macroscopically. In the distributed storage system as a whole, the probability of returning data from each node is β. Therefore, when there are n nodes in the distributed storage system, the ratio of returned packets is a value n × β obtained by multiplying the number n of nodes by the return probability β.
[0087]
The packet processing unit 272 combines the packets returned from each node. FIG. 13 is a diagram illustrating a data structure of a packet returned from the node. As shown in FIG. 13, the packet is composed of a header, a payload, and a footer, and the payload includes data extracted from the storage unit by each node, that is, FEC processing and interleave processing described above with reference to FIG. The header and the footer store control information such as a CRC checksum, a network address of a receiving node, and a sequence number indicating a packet order.
[0088]
When the reproduction instruction device receives the packet shown in FIG. 13, the packet processing unit 272 performs packet analysis, reads a sequence number, rearranges the order of the received packet, removes control information such as a header and a footer, and performs sequence analysis. Combine packets in numerical order.
[0089]
The data restoration processing unit 274 deinterleaves the received data blocks, arranges the data, and performs FEC decoding on the deinterleaved data to restore the original data.
[0090]
The data restoration processing based on the FEC and the deinterleave processing will be described with reference to FIG. As shown in FIG. 14A, a received packet from each node is lost on the network, and the received data block and the lost data block are mixed.
[0091]
As described above with reference to FIG. 5, the received data block is subjected to FEC encoding at an encoding rate of q / p, and the data divided into p blocks is converted into q encoded blocks. Block data.
[0092]
The data restoration processing unit 274 in the reproduction instruction device first deinterleaves the received data block, arranges the data arrangement, and generates deinterleaved processing data shown in FIG. The deinterleaved data includes a lost packet, that is, a data lost portion based on a lost block. However, due to the deinterleave processing, these errors are not burst errors in which an error portion exists as a large data area, but random errors including a small data area. Such a random error including a small data area can be eliminated by the FEC.
[0093]
The data restoration processing unit 274 performs error correction by FEC on the deinterleaved processing data shown in FIG. 14B, and generates restored data shown in FIG. 14C. As described above, FEC coding is a general term for coding methods that perform error correction on the receiving side, such as a tornado coding method, a lead tornado coding method, and a turbo coding method. When certain data is coded at the coding rate q / p, as described in the paper RIZZ097 (http://www.et.unipi.it/@luigi/fec.html#fec.ps), If more than one coded block remains, the original message can be restored even if some blocks are lost.
[0094]
The data restored by the data restoration processing unit 274 is output to the data processing unit 275. The data processing unit 275 stores the decoded data in a recording unit (not shown) or outputs the data to an output unit such as a monitor or a speaker via an output interface.
[0095]
The distributed storage system according to the present invention is a system in which data is recorded in each node with a recording probability α and data recorded in each node is returned with a return probability β, and is output from the recording instruction device 101 shown in FIG. The original data is returned at a rate of α × n × β. For example, if p blocks are coded into q blocks, q × α × n × β blocks are returned. As described in the above-mentioned article RIZZ097, if the number of returned blocks is larger than the number p of blocks before decoding, data can be decoded. Therefore, if the values of α, β, and q / p are determined so that the number of returned blocks is larger than p, target data can be decoded.
[0096]
That is, the rule determination condition setting unit 276 in the reproduction instruction device 270 determines that the reproduction target data stored in the node has a coding rate q / p obtained by performing FEC encoding on p blocks and converting the blocks into q blocks. When the data is coded processing data, a recording probability: α, a return probability: β and p, q specified by the recording instruction device connected to the network, the number of coded blocks: q, and a number of network connection nodes: n The relationship between the number of reply blocks that can be calculated by: q × α × n × β and the number of blocks: p is
Number of reply blocks: q × α × n × β> Number of blocks: p
Is set as β. With this setting, reliable data restoration from return data returned from each node with a return probability: β in response to a reproduction request from the reproduction instruction device is guaranteed.
[0097]
Thus, the distributed storage system in this example is
p ≧ q × α × n × β
Since the coding rate q / p, the recording probability α, and the return probability β may be set so as to satisfy the above, the coding rate q / p, the recording probability α, and the return probability β are changed within a range that satisfies the above-described formula. This makes it possible to change the data recording efficiency and transmission efficiency. Hereinafter, an example of setting each parameter of q / p, α, and β will be described.
[0098]
For example, when the value of the recording probability α is increased and the value of the response probability β is decreased for data having a very large number of reply requests, the data transmitted from each node decreases, and the data search processing at the node is performed. And data transmission processing is simplified.
[0099]
Also, instead of increasing the value of the recording probability α, the value of the coding rate q / p is increased and the response probability β is reduced, but the data search processing and data transmission processing at each node are simplified. can do.
[0100]
When the coding rate q / p is reduced and the recording probability α is increased, the number of packets to be transmitted can be suppressed. This is effective when p is large enough. Also, by reducing the recording probability α and decreasing the coding rate q / p, it is possible to avoid recording the same packet in a plurality of nodes. This is effective when p is sufficiently small.
[0101]
Also, by increasing the return probability β and decreasing the recording probability α or the coding rate q / p, the capacity of the coded data recorded in the entire distributed storage system can be reduced. Alternatively, when the probability of packet loss at the time of data recording, output, transmission, or the like is a, α, β, q / p such that a × n × α × q × β is sufficiently larger than p. , A sufficient number of data is returned.
[0102]
Also, by estimating the number of unique packets returned from a plurality of nodes mathematically and increasing the recording probability, return probability, or coding rate, the probability of unique packet arrival can be increased. .
[0103]
However, in the distributed storage processing configuration as described above, when data acquisition and reproduction are to be performed, packets storing the same data are received from a plurality of nodes. That is, a plurality of duplicate packets are received and stored in the buffer 273. As a result, there is a possibility that the available storage capacity of the buffer is significantly reduced due to useless data, and a case where effective packets cannot be efficiently received due to buffer overflow or the like occurs.
[0104]
Therefore, the return probability β set by the rule condition setting unit 276 shown in FIG. 11 is adjusted, and the transmission timing of the data request packet (see FIG. 12) is controlled by the phase control unit 278 to execute the control of the buffer 273. For efficient use, control for efficiently receiving valid data and storing it in the buffer 273 is executed.
[0105]
Due to the influence of network routing and the like, the time from transmission of a data request packet (see FIG. 12) as a playback command to arrival of a playback data storage packet (see FIG. 13) generally becomes a mountain-like graph like a normal distribution. . FIG. 15 is a graph showing a change in the number of input packets with respect to time for the playback instruction device when data is controlled to be input to the buffer of the playback instruction device at the same pace.
[0106]
The horizontal axis is time, and the vertical axis is a playback data storage packet (FIG. 13) that the playback instruction device (playback request node) receives from another node in response to transmission of a data request packet (see FIG. 12) as a playback command. ) Is the number of input packets. Each peak indicates a received packet for a different data request. As can be understood from the figure, the number of input packets for each request increases with the passage of time, decreases after forming peaks (peaks of each mountain). The data shown in FIG. 15 is based on the assumption that the above-mentioned return probability β is set to be constant in each request (data request packet). Has a similar transition.
[0107]
Next, transition of a response packet (reproduced data storage packet) for each request in the case where the return probability β set in the data request packet by the reproduction instruction device is changed will be described with reference to FIG.
[0108]
In FIG. 16, (a) shows the change in the number of received packets of the response packet (reproduced data storage packet) when the return probability β is set to a high value, and (b) shows the case when the return probability β is set to a low value. I have.
[0109]
If the return probability β is set to a high value, the processing probability at the node that has received the data request packet, that is, the request data is obtained from its own storage means, and a reproduction data storage packet (see FIG. 13) is generated. 16A, the transmission probability increases, and as shown in FIG. On the other hand, if the reply probability β is set low, the generation and reply probability of the reproduced data storage packet at the node that has received the data request packet is reduced, and as shown in FIG. It will show the transition.
[0110]
On the reproduction instructing device side, the larger the return probability β is, the faster the valid packets that can fill the buffer are obtained. However, for one request, as described above, for example, p blocks are used. In the data coded into q blocks, q × α × n × β blocks corresponding to the setting of the return probability β are returned. As described in the paper RIZZ097, when the number of returned blocks is larger than the number p of blocks before decoding, data can be decoded. Therefore, if the number of returned blocks reaches p, the target data can be decoded. Therefore, packets exceeding p packets are wasted packets. Storing these useless packets in the buffer only reduces the remaining capacity of the buffer that can be stored, and hinders securing a storage area for response packets to new requests.
[0111]
Conversely, when the return probability β is set low as shown in FIG. 16B, the number of packets that are delayed in time and wasted decreases, but the speed of filling the buffer decreases.
[0112]
In view of the above situation, in the reproduction instruction device (reproduction request node) according to the present invention, the phase control unit 278 shown in FIG. 11 controls the transmission timing of the data request packet as the reproduction command, and sets the rule determination condition setting unit 276. , The value of the return probability β set in the data request packet is adjusted. By these controls, efficient use of the buffer and reduction of reception of useless packets are realized.
[0113]
FIG. 17 is a diagram illustrating a control example in the reproduction instruction device (reproduction request node) of the present invention. As in FIGS. 15 and 16, the horizontal axis represents time, and the vertical axis represents the reproduction instruction device (reproduction request node) received from another node as a response to the transmission of a data request packet (see FIG. 12) as a reproduction command. This is the number of input packets of reproduced data storage packets (see FIG. 13) to be reproduced. Each peak indicates a received packet for a different data request.
[0114]
FIG. 17 shows the transition of the number of input packets of reproduced data storage packets corresponding to four different continuous data request requests (a) to (d). When transmitting a continuous data request, the value of the return probability β set in the data request packet transmitted at the beginning of the request is set to a high value, and the setting value of the return probability β is adjusted to decrease as time passes.
[0115]
As described above, the reproduction target data stored in the node has a coding rate q / p obtained by dividing the data into p blocks and subjecting the p blocks to FEC encoding to convert the blocks into q blocks. The rule determination condition setting unit is encoding processing data, and is configured to set, as a reproduction rule determination condition description, a probability value at which data is returned with a return probability β at a node that receives a data reproduction processing request, and is connected to a network. The relationship between the recording probability α specified by the recording instruction device and the number of coded blocks q, the number of reply blocks q × α × n × β that can be calculated from the number n of network connection nodes, and the number of blocks p is as follows: Number of reply blocks: q × α × n × β> Number of blocks: p Return probability: β is used as a reference value, and when the remaining buffer capacity is large, it is set in a data reproduction processing request packet. It sets high value of signal probability beta, when the buffer remaining capacity is small performs control to set a low value of the return probability beta.
[0116]
That is, at first, a high return probability β value is set that allows more wasteful packets to fill the buffer earlier. That is, since the buffer capacity is sufficient, high-speed reception is prioritized. As the buffer free space gradually decreases, a low return probability β is set in order to prevent reception of useless packets. That is, processing is performed in consideration of buffer capacity rather than high-speed reception.
[0117]
However, a decrease in the reception speed causes a decrease in the reproduction efficiency, which is a problem. Therefore, in order to prevent the reception speed from decreasing, the transmission timing of the data request packet from the reproduction instruction device is advanced.
[0118]
For example, in data in which p blocks are coded into q blocks, in a configuration in which q × α × n × β blocks are returned according to the setting of the return probability β, the number of return blocks is If the number of returned blocks is larger than p, the data can be decoded. When the number of returned blocks reaches almost p, the data request is received so that a response packet to the next request is received. Set the packet transmission timing.
[0119]
Times T1, T2, and T3 shown in FIG. 17 indicate reception intervals of response packets corresponding to the data request packets. However, the transmission timing of the data request packet in the reproduction instruction device is also the time T1, T2, and T3 shown in FIG. Is set to the interval corresponding to. That is, the time until the next reproduction command (data request packet) is sent gradually increases.
[0120]
As described above, the rule determination condition setting unit 276 in the reproduction instruction device shown in FIG. 11 sets the initial value high in the process of setting the value of the return probability β set in the reproduction condition rule of the data request packet as the reproduction command. In the subsequent request, control to gradually lower the set value is performed. In addition, the phase control unit 278 executes control to shorten the initial request interval and gradually increase the request transmission interval, such as the times T1, T2, and T3 shown in FIG. 17, for the transmission timing of the data request packet.
[0121]
These controls can be performed more reliably by controlling based on the number of packets arriving at the buffer 273 and the remaining capacity information of the buffer 273.
[0122]
A description will be given of a reproduction processing procedure in which a node that has received a packet autonomously determines whether or not to extract and transmit content based on transmission of a reproduction processing request packet from the reproduction instruction device and executes the processing.
[0123]
FIG. 18 is a diagram illustrating a process of transmitting the data reproduction process request packet shown in FIG. 12 as a data reproduction command from the reproduction instruction device 102 to the plurality of nodes 111 to 115. The reproduction instruction device sets an identifier of the reproduction designation data, for example, a content ID, a GUID, or the like as a request description, and further generates a data reproduction processing request packet shown in FIG. 12 in which a probability value: β is set as a rule determination condition description. Then, the packet is unicast or multicast delivered to each node.
[0124]
FIG. 19 shows whether or not the node receiving the data reproduction request packet shown in FIG. 12 executes the data reproduction processing, that is, reads the designated data from the storage unit of its own node, generates the packet, and transmits the packet according to the reproduction command. A packet in which reproduced data is stored only at the node that has determined autonomously that the processing has been performed, that is, a packet having processed data subjected to the FEC processing and interleaving processing described in FIG. 6 as a payload (see FIG. 13) ) Is generated and transmitted to the reproduction instruction device 112.
[0125]
Referring to FIG. 20, a processing procedure in the node that has received the data reproduction processing request packet shown in FIG. 12 in which the probability value: β is set as the rule determination condition description will be described. First, in step S301, a data reproduction processing request packet is awaited. If it is determined in step S302 that the packet has been received, a random number generation process is executed in step S303. In step S304, the generated random number and the data reproduction processing request Description of playback rule determination condition stored in packet: comparison with probability β, and execution / non-execution of instruction is determined based on the comparison result.
[0126]
For example, if the generated random number> probability β, the designated content stored in its own storage means is extracted, a packet having the extracted data as a payload (see FIG. 13) is generated, and transmitted to the reproduction instruction device. The process is executed, and if the generated random number ≦ probability β, the data reproduction process is not executed.
[0127]
If it is determined that the data reproduction processing request should be executed as a result of the comparison processing based on the random number in step S304, the process proceeds to step S305, and the processing of extracting the processed data stored in the storage unit of the own device (node) is performed. Then, a process of generating a packet with the extracted data as a payload (see FIG. 13) and transmitting the packet to the reproduction instruction device is performed. On the other hand, if it is determined that the data reproduction processing request is not to be executed as the comparison processing result based on the random number, the process proceeds to step S306, and the processing ends without executing the data reproduction processing.
[0128]
As described above, the information processing apparatus that has received the data reproduction processing request (instruction) autonomously determines the execution or non-execution of the instruction based on the comparison between the reproduction rule determination condition description stored in the packet and the random number. , And performs processing according to the determination.
[0129]
Next, with reference to the processing flow of FIG. 21, the information processing apparatus that has received the data reproduction processing request calculates a hash value, and autonomously executes and does not execute the data reproduction processing request based on the hash value. Will be described.
[0130]
First, in step S401, a data reproduction request packet is awaited. If it is determined in step S402 that the packet has been received, a hash value generation process based on data stored in the received packet is executed in step S403, and in step S404, An execution or non-execution of an instruction is determined based on the generated hash value.
[0131]
For example, the information processing apparatus stores a preset value as a preset threshold value in the storage unit, and executes the command if the generated hash value is greater than the set value, and executes the instruction if the generated hash value is smaller than or equal to the set value. Make settings such as execution.
[0132]
The data for which the hash value is to be generated is, for example, an identifier (data ID) of the content data or a part of the data, for example, X bits from the beginning of the data shown in FIG. It is possible to set data as hash target data. For example, MD5 can be applied to the hash value calculation, and a hash value based on MD5 of the data ID or a hash value based on MD5 of the data content is generated as shown in FIG.
[0133]
If it is determined that the data recording processing request should be executed as a result of the determination based on the hash value in step S404, the process proceeds to step S405, and the processing of extracting the processed data stored in the storage unit of the own device (node) is performed. Then, a process of generating a packet with the extracted data as a payload (see FIG. 13) and transmitting the packet to the reproduction instruction device is performed. On the other hand, if it is determined that the data reproduction processing request is not to be executed as the comparison processing result based on the hash value, the process proceeds to step S406, and the processing ends without executing the data reproduction processing.
[0134]
As described above, the information processing apparatus that has received the data reproduction processing request (command) autonomously executes and non-executes the command based on the comparison between the hash value of the data stored in the packet and the set value of each node. Is determined, and processing according to the determination is performed.
[0135]
[5. Data recording / playback processing sequence]
Next, referring to FIGS. 22 and 23, the recording instruction device issues a data recording processing request to the node, the node performs data recording processing, the reproduction instruction device issues a data reproduction processing request to the node, the node extracts data, A series of processing sequences of the transmission processing will be described together.
[0136]
In FIG. 22, the recording instruction apparatus first performs processing of recording target data (contents), that is, FEC processing and interleave processing in step S11. This process is the process described above with reference to FIG.
[0137]
Next, in step S12, the recording instruction device executes a process of generating a data recording process request (command) packet. The data subjected to the interleave processing is stored as a payload, and a packet is generated in which a probability value: α is set as a [rule determination condition description] determined by the rule determination condition setting unit 253 (see FIG. 4).
[0138]
Next, in step S13, the recording instruction device transmits an address to each node configuring the distributed storage system using an address setting corresponding to the transmission node, that is, using unicast or multicast.
[0139]
The processing of the node which has received the data recording processing request (command) packet from the recording instruction device is performed by autonomously determining whether to execute the processing described with reference to FIGS. Processing. FIG. 22 shows the processing of two nodes (nodes 1 and 2). In many other nodes, an autonomous judgment (probability control) is executed and a node that executes data recording processing There are nodes that do not execute the data recording process.
[0140]
The two nodes (nodes 1 and 2) shown in FIG. 22 perform the probability control processing of steps S21 and S31, that is, the generation random number described with reference to FIG. 8 and the recording in the received packet (data recording processing request packet). Probability set as rule determination condition description: processing execution / non-execution determination processing based on comparison with α, or a hash value based on data in a packet described with reference to FIG. Is executed based on the comparison of the above.
[0141]
As a result of the probability control processes in steps S21 and S31, it is concluded that the two nodes (nodes 1 and 2) shown in FIG. 22 execute the data recording process. In steps S22 and S32, the data recording process is performed. Execute The recording target data is the FEC and the interleaved processed data stored in the data recording processing request packet received from the recording instruction device.
[0142]
Next, a reproduction processing sequence will be described with reference to FIG. In step S41, the reproduction instruction device transmits a reproduction processing request (command) packet (see FIG. 12) to each node. This is a packet in which a content ID or the like to be reproduced is stored as a request description and a reproduction rule determination condition description (probability: β) is set. The value of the return probability β as the description of the reproduction rule determination condition is initially set to a relatively high value as described above.
[0143]
Each node that has received the reproduction processing request (command) packet (see FIG. 12), that is, the two nodes (nodes 1 and 2) shown in FIG. 23, refers to the probability control processing of steps S51 and S61, that is, to FIG. The processing execution / non-execution determination processing based on a comparison between the generated random number described above and the probability: β set as the reproduction rule determination condition description in the received packet (data reproduction processing request packet), or FIG. The process execution / non-execution determination process is performed based on a comparison between the hash value based on the in-packet data described with reference to the node setting value.
[0144]
As a result of the probability control process in step S51, the conclusion that the node 1 shown in FIG. 23 executes the data reproduction process is obtained. In step S52, the node 1 stores the data in its own node according to the request description of the reproduction process request (command) packet. The corresponding data is obtained from the unit, and a packet (see FIG. 13) storing the obtained data as a payload is generated and transmitted to the reproduction instruction device in step S53.
[0145]
On the other hand, as a result of the probability control process in step S61, the node 2 shown in FIG. 23 determines that the data reproduction process is not to be executed, and ends the process without executing data extraction, packet generation, and transmission.
[0146]
Although FIG. 23 shows a configuration in which data only from node 1 is transmitted to the reproduction instruction device, data storage packets from nodes 3 to n (not shown) are transmitted to the reproduction instruction device, The device is receiving a large number of packets from a large number of nodes.
[0147]
In step S71, the reproduction instruction device that has received the packet storing the reproduction data applies deinterleaving to the received data block, arranges the data arrangement, further performs FEC decoding on the deinterleaved data, and converts the original data. Restore.
[0148]
The data restoration processing based on the FEC and the deinterleave processing is as described above with reference to FIG. Even if there is a lost packet, the error becomes a random error instead of a burst error in which an error part exists as a large data area by the deinterleave processing, and the original data is restored by the FEC decoding processing.
[0149]
Next, in step S72, the reproduction instruction device sets the value of the return probability β to be set in the reproduction condition rule of the data request packet as the reproduction command to be transmitted next, and sets the transmission timing of the data request packet.
[0150]
In setting the value of the reply probability β, the rule determination condition setting unit 276 shown in FIG. 11 performs control to set the initial value higher as described above and gradually lower the set value in the subsequent request. Also, the phase control unit 278 shortens the initial request interval as in the time T1, T2, and T3 as described earlier with reference to FIG. Execute the control to lengthen. These controls are performed based on the number of packets arriving at the buffer 273 and the remaining capacity information of the buffer 273.
[0151]
In the sequence diagram shown in FIG. 23, the process of setting the return probability β and the phase as the sending timing in step S72 is shown as a sequence in which a response packet to a previous request is received and a deinterleave process to the response packet is performed. However, the process of setting the return probability β and the phase as the sending timing may be performed before receiving a response packet to the previous request. That is, the process of step S72 is executed based on the number of packets arriving at the buffer 273 and the remaining capacity information of the buffer 273, and the next request is transmitted.
[0152]
In step S73, the reproduction instruction device transmits a reproduction processing request (command) packet (see FIG. 12) in which the return probability β set in step S72 is set to each node based on the dispatch timing set in step S72. I do.
[0153]
Each node that has received the reproduction processing request (command) packet (see FIG. 12), that is, the two nodes (nodes 1 and 2) shown in FIG. 23, refers to the probability control processing of steps S81 and S82, that is, to FIG. The processing execution / non-execution determination processing based on a comparison between the generated random number described above and the probability: β set as the reproduction rule determination condition description in the received packet (data reproduction processing request packet), or FIG. The process execution / non-execution determination process is performed based on a comparison between the hash value based on the in-packet data described with reference to the node setting value.
[0154]
As a result of the probability control process in step S82, the conclusion that the node 1 shown in FIG. 23 executes the data reproduction process is obtained. In step S83, the node 1 stores the data in its own node according to the request description of the reproduction process request (instruction) packet. The corresponding data is obtained from the unit, and a data storage packet (see FIG. 13) storing the obtained data as a payload is generated, and is transmitted to the reproduction instruction device in step S84.
[0155]
On the other hand, as a result of the probability control process in step S81, the node 2 illustrated in FIG. 23 determines that the data reproduction process is not to be performed, and ends the process without performing data extraction, packet generation, and transmission.
[0156]
In step S91, the playback instruction device that has received the packet storing the playback data performs deinterleaving on the received data block, arranges the data arrangement, further performs FEC decoding on the deinterleaved data, and converts the original data. Restore.
[0157]
If there is more data to be requested, the reproduction support apparatus sets the return probability β and sets the phase in the same manner as in step S72, and repeats the data request packet transmission process based on the setting.
[0158]
As described above, the reproduction instruction device that executes the data reproduction request controls the value of the return probability β set for the data request packet based on the number of packets arriving at the buffer and the remaining capacity of the buffer, and the transmission timing of the data request packet. . If the remaining buffer capacity is large, the value of the return probability β is set high, and if the remaining buffer capacity is small, control is performed to decrease the value of the return probability β. The phase control for transmitting the next request is executed on condition that the number reaches the number p or on the condition that the number approaches the number p.
[0159]
The return probability β and the phase control process realize efficient use of a buffer and high-speed reception of data.
[0160]
Next, a processing procedure in the reproduction instruction device will be described with reference to FIG. First, in step S501, the reproduction instruction device transmits a reproduction processing request (command) packet (see FIG. 12) to each node. This is a packet in which a content ID or the like to be reproduced is stored as a request description and a reproduction rule determination condition description (probability: β) is set. The value of the return probability β as the description of the reproduction rule determination condition is initially set to a relatively high value as described above.
[0161]
The reproduction instruction device waits for the reproduction data storage packet in step S502, and when reception is confirmed in step S503, in step S504, the reproduction instruction device deinterleaves the received data block, arranges the data arrangement, and further deinterleaves. FEC decoding is performed on the obtained data to restore the original data.
[0162]
Next, in step S505, the reproduction instruction device determines whether or not data reception regarding the reproduction processing content has been completed. If it is determined that data reception has not been completed, the return probability set in the next data request packet in step S506 The value of β and the setting of the packet transmission timing are set based on the number of packets arriving in the buffer and the remaining buffer capacity, and in step S507, after waiting for the set packet transmission timing, the data request packet is transmitted in step S501.
[0163]
[6. Synchronous processing during recording and playback]
Next, a description will be given of a synchronous processing configuration in which data recording processing by a recording instruction device for a plurality of nodes on a network and reproduction processing by a reproduction instruction device are executed in parallel.
[0164]
FIG. 25 shows that the recording instruction device 101 transmits a data recording instruction to the distributed nodes 111 to 115, and each node executes data recording based on the autonomous judgment described above. A configuration example in which the device 102 executes data reproduction based on the transmission of the data request packet for data being recorded is shown.
[0165]
In this parallel processing configuration, if the pace at which the recording instructing device 101 performs data recording and the pace at which the playback instructing device 102 performs data playback are different, a buffer underflow or overflow occurs in the playback instructing device 102, and a playback processing error occurs. May occur. In order to prevent such a reproduction processing error, a timing generation device 311 is configured in the recording instruction device 101, and a timing synchronization device 312 is configured in the reproduction instruction device 102.
[0166]
The timing generation device 311 of the recording instruction device 101 transmits a timing control packet storing transmission timing information of the recording instruction transmitted to each of the nodes 111 to 115 as, for example, time stamp data to the timing synchronization device 312 of the reproduction instruction device 102. I do.
[0167]
The reproduction instruction device 102 extracts the timing control packet received by the timing synchronization device 312, acquires data recording processing timing information, and transmits a reproduction command to each of the nodes 111 to 115 based on the acquired information. It also controls the timing of extracting data from the buffer.
[0168]
That is, the reproduction instruction device 102 transmits a data request packet as a process of acquiring recording target data to the nodes 111 to 115 after a certain time after the timing at which the recording instruction device 101 executes data recording. The time after a certain time is a time in consideration of the time when each node receives the recording command from the recording instruction device 101 and executes the data recording process. By performing the timing control in this manner, each of the nodes 111 to 115 receives a data reproduction request from the reproduction instruction device 102 after performing data recording based on a recording command from the recording instruction device. Thus, the stored data can be taken out and transmitted, and the reproduction instruction device 102 can reliably receive the data and execute the reproduction.
[0169]
FIG. 26 shows a configuration example in which the timing generator 321 is set independently and separately, and the timing synchronizers 322 and 323 are set on the recording instruction device 101 side and the reproduction instruction device 102 side, respectively.
[0170]
The timing generation device 321 transmits a timing control packet storing timing information, for example, as time stamp data to the timing synchronization devices 322 and 323 on the recording instruction device 101 side and the reproduction instruction device 102 side.
[0171]
The recording instruction device 101 and the reproduction instruction device 102 take out the timing control packet received from the timing generation device 321 and control the data recording or reproduction processing timing. That is, under the timing control based on the acquired timing information, a recording command or a reproduction command is transmitted to each of the nodes 111 to 115, and the timing of extracting data from the buffer is controlled. By performing the processing in synchronization with each other, the reproduction processing can be performed in accordance with the progress of the recording processing without causing any delay in the recording processing or the reproduction processing.
[0172]
FIG. 27 shows a sequence diagram in which the recording instruction device and the reproduction instruction device respectively execute the issuance control of the recording instruction and the reproduction instruction based on the timing control information output from the independently provided timing generating device.
[0173]
In step S700, the timing generation device transmits a timing control packet in which a time stamp is set based on its own clock information to the recording instruction device and the reproduction instruction device. After transmitting the timing control packet in step S700, the timing generation device periodically repeats the timing control packet in which the time stamp is set, and transmits it to the recording instruction device and the reproduction instruction device.
[0174]
The recording instruction device executes a process of transmitting a recording command packet to the node based on the time stamp information stored in the timing control packet received from the timing generation device and its own clock information. The processing in steps S701 to S703 is the same as the processing described above with reference to FIG. 22, and a description thereof will be omitted.
[0175]
Hereinafter, the recording instruction device controls the transmission timing of the recording command packet to the node based on the timing control packet continuously received from the timing generation device, and executes the packet transmission process.
[0176]
On the other hand, the reproduction instructing device executes transmission processing of a reproduction instruction packet to the node based on the time stamp information stored in the timing control packet received from the timing generating device and its own clock information. The processing in steps S704 to S708 is the same as the processing described above with reference to FIG. 23, and a description thereof will not be repeated.
[0177]
Hereinafter, the reproduction instruction device controls the transmission timing of the reproduction command packet to the node based on the timing control packet continuously received from the timing generation device, and executes the packet transmission process.
[0178]
As described above, the recording instruction device and the reproduction instruction device can control the transmission timing of the recording instruction and the reproduction instruction based on the common timing control packet received from the timing generation device, and perform the synchronized processing. This makes it possible to perform a reproduction process in accordance with the progress of the recording process without causing any delay in the recording process or the reproduction process.
[0179]
[7. Hardware configuration of information processing device]
Next, an example of a hardware configuration of the information processing device that configures the node, the recording instruction device, the reproduction instruction device, and the timing generation device described in the above embodiment will be described.
[0180]
FIG. 28 illustrates an example of an information processing apparatus including a CPU (Central Processing Unit) as a control unit. The configuration shown in FIG. 28 will be described. A CPU (Central Processing Unit) 901 is a processor that executes various programs. A ROM (Read-Only-Memory) 902 stores a program executed by the CPU 901 or fixed data as operation parameters. A RAM (Random Access Memory) 903 is used as a storage area and a work area for a program executed in the processing of the CPU 901 and parameters that change as appropriate in the program processing.
[0181]
The HDD 904 performs control of the hard disk, and stores and reads various data and programs from and to the hard disk. The encoding / decoding processing unit 905 performs the encoding processing of transmission data such as content and the decoding processing of reception data in accordance with the above-described processing.
[0182]
The bus 921 is configured by a PCI (Peripheral Component Internet / Interface) bus or the like, and enables data transfer with each module and each input / output device via the input / output interface 922.
[0183]
The input unit 911 is an input unit that includes, for example, a keyboard and a pointing device. When the input unit 911 is operated via a keyboard, a mouse, or the like, or when data is received from the communication unit 913, a command is input to the CPU 901 and a program stored in a ROM (Read Only Memory) 902 is stored. Execute The output unit 912 is, for example, a CRT, a liquid crystal display, or the like, and displays various types of information as text or images.
[0184]
The communication unit 913 executes communication between information processing apparatuses or communication processing with other entities, and under the control of the CPU 901, data supplied from each storage unit, or the CPU 901 and the encode / decode processing unit 905. A process for transmitting processed data and receiving data from another entity is executed.
[0185]
The drive 914 is a drive that performs recording and reproduction of a removable recording medium 915 such as a flexible disk, a CD-ROM (Compact Disc Only Memory), an MO (Magneto optical) disk, a DVD (Digital Versatile Disc), a magnetic disk, and a semiconductor memory. Yes, it executes the reproduction of the program or data from each removable recording medium 915 and the storage of the program or data in the removable recording medium 915.
[0186]
When a program or data recorded in each storage medium is read and executed or processed in the CPU 901, the read program and data are supplied to, for example, the connected RAM 903 via the input / output interface 922 and the bus 921.
[0187]
The processes described in the specification can be executed by hardware, software, or a combination of both. When a series of processing is executed by software, it is possible to execute various functions by installing a computer in which a program constituting the software is incorporated in dedicated hardware or various programs, For example, the program may be stored in a program readable recording medium such as a flexible disk or a CD-ROM and provided to a general-purpose personal computer, or the program may be downloaded via a communication network such as the Internet. .
[0188]
Specifically, the program can be recorded in a hard disk or a ROM (Read Only Memory) as a recording medium in advance. Alternatively, the program is temporarily or permanently stored on a removable recording medium such as a flexible disk, a CD-ROM (Compact Disc Only Memory), an MO (Magneto optical) disk, a DVD (Digital Versatile Disc), a magnetic disk, or a semiconductor memory. It can be stored (recorded). Such a removable recording medium can be provided as so-called package software.
[0189]
In addition to installing the program on the computer from the above-described removable recording medium, the program can be wirelessly transferred from the download site to the computer, or transferred to the computer via a network such as a LAN (Local Area Network) or the Internet by wire. The computer can receive the program transferred in this way and install it on a recording medium such as a built-in hard disk.
[0190]
As the information processing device constituting the node, the recording instruction device, and the reproduction instruction device, a camcorder, a personal video recorder, a home gateway, and the like can be considered. Alternatively, the apparatus may have another configuration as long as it has a network interface for transmitting and receiving data.
[0191]
Further, although the recording probability α and the reply probability β are recorded in the packet, the recording probability α and the reply probability β may be recorded in any recording device or packet, and each node may refer to the values. . When the read tornado coding method is used as the FEC coding, the interleaving process can be omitted.
[0192]
The various processes described in the specification may be executed not only in chronological order according to the description but also in parallel or individually according to the processing capability of the device that executes the processes or as necessary. Further, in this specification, a system is a logical set configuration of a plurality of devices, and is not limited to a device having each configuration in the same housing.
[0193]
The present invention has been described in detail with reference to the specific embodiments. However, it is obvious that those skilled in the art can modify or substitute the embodiments without departing from the spirit of the present invention. That is, the present invention has been disclosed by way of example, and should not be construed as limiting. In order to determine the gist of the present invention, the claims described at the beginning should be considered.
[0194]
【The invention's effect】
As described above, according to the configuration of the present invention, in the reproduction instruction apparatus that transmits a data reproduction processing request to a node connected to the network, the receiving node of the reproduction processing request packet performs processing in accordance with the processing request. Is changed according to the packet transmission sequence, and the transmission timing of the data reproduction processing request packet is controlled. Set a high return probability β value that allows more packets, and when the buffer free space decreases, execute processing to set a low return probability β to prevent the reception of useless packets, and transmit the packet. By adjusting the interval, it is possible to perform data reproduction that achieves efficient use of the buffer and high-speed reception. .
[0195]
Further, according to the configuration of the present invention, the recording probability α specified by the recording instruction device connected to the network, the number of coded blocks q, and the number of return blocks q × α that can be calculated from the number of network connection nodes n When the relationship between × n × β and the number of blocks p is the number of reply blocks: q × α × n × β> the number of blocks: p, the return probability: β is used as a reference value, and when the remaining buffer capacity is large, The value of the reply probability β set in the data reproduction processing request packet is set to a high value, and if the remaining buffer capacity is small, the control is performed to set the value of the reply probability β to a low value. As a result, the buffer use efficiency is improved, and high-speed reception is realized.
[0196]
Further, according to the configuration of the present invention, it becomes possible for the reproduction instruction device to control the transmission timing of the reproduction command based on the timing control packet received from the timing generation device, and execute the recording process and the reproduction process in parallel. In this case, the reproduction process can be performed in accordance with the progress of the recording process without causing a recording delay or a reproduction process delay.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a configuration example of a distributed storage system to which the present invention can be applied.
FIG. 2 is a diagram illustrating an example of a network configuration to which the present invention can be applied.
FIG. 3 is a diagram illustrating a configuration of an information processing apparatus (node) connected to a network.
FIG. 4 is a diagram illustrating a configuration of a recording instruction device.
FIG. 5 is a diagram illustrating FEC encoding processing and interleaving processing of data executed in the recording instruction device.
FIG. 6 is a diagram illustrating a configuration of a data recording processing request packet transmitted from the recording instruction device.
FIG. 7 is a diagram illustrating transmission processing of a data recording processing request packet from a recording instruction device to a node.
FIG. 8 is a flowchart illustrating a processing procedure including an autonomous processing execution determination process in a node that has received a data recording processing request packet from the recording instruction device.
FIG. 9 is a flowchart illustrating a processing procedure including an autonomous processing execution determination processing in a node that has received a data recording processing request packet from the recording instruction device.
FIG. 10 is a diagram illustrating an example of a hash value generated in an autonomous process execution determination in a node that has received a data recording process request packet from a recording instruction device.
FIG. 11 is a diagram illustrating a configuration of a reproduction instruction device.
FIG. 12 is a diagram illustrating a configuration of a data reproduction processing request packet transmitted from the reproduction instruction device.
FIG. 13 is a diagram illustrating a configuration of a packet transmitted from the node that has received the data reproduction processing request packet to the reproduction instruction device.
FIG. 14 is a diagram for describing data deinterleaving processing and FEC decoding processing executed in the reproduction instruction device.
FIG. 15 is a diagram illustrating an example of transition data of the number of packets received by the reproduction instruction device.
FIG. 16 is a diagram illustrating the correspondence between the transition of the number of packets received by the reproduction instruction device and the set value of the return probability β.
FIG. 17 is a diagram showing an example of transition data of the number of received packets based on the return probability and the phase control of the reproduction instruction device of the present invention.
FIG. 18 is a diagram illustrating transmission processing of a data reproduction processing request packet from a reproduction instruction device to a node.
FIG. 19 is a diagram illustrating transmission processing of a data storage packet from a node that has received a data reproduction processing request packet from the reproduction instruction device.
FIG. 20 is a flowchart illustrating a processing procedure including an autonomous processing execution determination processing in a node that has received a data reproduction processing request packet from the reproduction instruction device.
FIG. 21 is a flowchart illustrating a processing procedure including an autonomous processing execution determination process in a node that has received a data reproduction processing request packet from a reproduction instruction device.
FIG. 22 is a sequence diagram illustrating a data recording process request from a recording instruction device and a data recording process in a node.
FIG. 23 is a sequence diagram illustrating a data reproduction processing request from a reproduction instruction device, a data extraction and transmission process in a node, and a reproduction process in a reproduction instruction device.
FIG. 24 is a flowchart illustrating a processing procedure in the playback instruction device.
FIG. 25 is a diagram illustrating a synchronous processing configuration of a recording / reproducing process based on a timing generation device and a synchronization device.
FIG. 26 is a diagram illustrating a synchronous processing configuration of a recording / reproducing process based on a timing generator and a synchronizer provided independently.
FIG. 27 is a diagram illustrating a recording / reproducing processing sequence based on timing control information from a timing generator.
FIG. 28 is a diagram illustrating an example of a hardware configuration of an information processing apparatus according to the present invention.
[Explanation of symbols]
100 networks
101 Recording instruction device
102 Playback instruction device
111-115 nodes
110 network
121-126 Information processing device
200 Information processing device (node)
201 Rule judgment processing unit
202 Data processing unit
203 Packet processing unit
204 Data transmission / reception unit
250 Recording instruction device
251 data input section
252 Data processing unit
253 Rule judgment condition setting part
254 packet generator
255 network interface
270 Playback instruction device
271 Network Interface
272 Packet processing unit
273 buffer
274 Data restoration processing unit
275 Data processing unit
276 Rule judgment condition setting section
277 Data request packet generator
278 Phase control unit
311 Timing generator
312 Timing synchronizer
321 Timing generator
322,323 timing synchronizer
901 CPU
902 ROM
903 RAM
904 HDD
905 Encoding / decoding processing unit
911 input section
912 output unit
913 Communication unit
914 drive
915 Removable storage medium
921 bus
922 I / O interface

Claims (15)

An information processing device as a reproduction instruction device for transmitting a data reproduction processing request to a node connected to a network,
A packet generation unit that generates a data reproduction processing request packet;
A rule for changing and setting the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence A judgment condition setting unit;
A phase control unit that controls a transmission timing of the data reproduction processing request packet, a network interface unit that transmits the packet generated by the packet generation unit,
A buffer for storing packet storage data received from a node as a response to the data reproduction processing request packet. The rule determination condition setting unit,
2. The information processing apparatus according to claim 1, wherein in the data reproduction processing request packet transmission sequence to be continuously transmitted, processing is performed to set the value of the return probability β as a gradually reduced value. . The phase control unit,
2. The information processing apparatus according to claim 1, wherein a control process for setting a data reproduction process request packet transmission interval to be gradually longer in a data reproduction process request packet transmission sequence to be continuously transmitted is performed. The data to be reproduced stored in the node is obtained by dividing the data into p blocks and subjecting the p blocks to FEC coding to convert the blocks into q blocks, thereby performing an encoding process at a coding rate q / p. Data,
The rule determination condition setting unit,
A configuration in which a probability value for returning data with a return probability β at a node receiving the data reproduction processing request is set as a reproduction rule determination condition description, and a recording probability α specified from a recording instruction device connected to the network, And the relationship between the number of coded blocks q, the number of reply blocks q × α × n × β that can be calculated from the number n of network connection nodes, and the number of blocks p,
Number of reply blocks: q × α × n × β> Number of blocks: p
Return probability: β is used as a reference value. When the remaining buffer capacity is large, the value of the return probability β set in the data reproduction processing request packet is set high, and when the remaining buffer capacity is small, the return probability β is set. The information processing apparatus according to claim 1, wherein the information processing apparatus is configured to perform control for setting a value to be low. The phase control unit,
On condition that the reception response packet to the data reproduction processing request as the previous request has reached the value p, which is the number of packets that can be decoded, or on condition that the number is close to the value p, the following The information processing apparatus according to claim 4, wherein the information processing apparatus is configured to execute phase control for transmitting a request. The information processing device further includes:
A data restoration processing unit that performs deinterleaving processing and FEC decoding processing;
The data restoration processing unit is configured to execute deinterleaving processing and FEC decoding processing on reproduction target data extracted from a packet received from a node that has received the data reproduction processing request, and perform data restoration. The information processing apparatus according to claim 1. The information processing device further includes:
The information processing apparatus according to claim 1, further comprising a timing synchronization processing unit that receives synchronization control information from outside and controls transmission timing of the data reproduction processing request packet. An information processing method for performing data reproduction processing control based on transmission of a data reproduction processing request to a node connected to a network,
Rule determination for changing the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence A condition setting step;
A packet generation step of generating a data reproduction processing request packet;
A phase control step of controlling a transmission timing of the data reproduction processing request packet,
An information processing method comprising: The rule determination condition setting step includes:
9. The information processing method according to claim 8, wherein, in a data reproduction processing request packet transmission sequence to be continuously transmitted, a process of setting the value of the return probability β as a gradually reduced value is executed. The phase control step includes:
9. The information processing method according to claim 8, wherein in the data reproduction processing request packet transmission sequence to be continuously transmitted, control processing for setting a data reproduction processing request packet transmission interval to be gradually longer is executed. The data to be reproduced stored in the node is obtained by dividing the data into p blocks and subjecting the p blocks to FEC coding to convert the blocks into q blocks, thereby performing an encoding process at a coding rate q / p. Data,
The rule determination condition setting step includes:
A probability value for returning data with a return probability β at a node receiving the data reproduction processing request is set as a reproduction rule determination condition description, and a recording probability α specified by a recording instruction device connected to the network and the code The relationship between the number of coded blocks q, the number of reply blocks q × α × n × β that can be calculated from the number of network connection nodes n, and the number of blocks p is as follows:
Number of reply blocks: q × α × n × β> Number of blocks: p
Return probability: β is used as a reference value. When the remaining buffer capacity is large, the value of the return probability β set in the data reproduction processing request packet is set high, and when the remaining buffer capacity is small, the return probability β is set. 9. The information processing method according to claim 8, wherein control is performed to set the value low. The phase control step includes:
On the condition that the reception response packet to the data reproduction processing request as the previous request has reached the value p, which is the number of packets that can be decoded, or on condition that the number is close to the value p, the following The information processing method according to claim 11, wherein phase control for transmitting the request is performed. The information processing method further includes:
Having a data restoration processing step of performing deinterleaving processing and FEC decoding processing;
The data restoration processing step performs data restoration by executing deinterleaving processing and FEC decoding processing on reproduction target data extracted from a packet received from a node that has received the data reproduction processing request. Item 9. The information processing method according to Item 8. The information processing method further includes:
9. The information processing method according to claim 8, further comprising a timing synchronization processing step of receiving synchronization control information from outside and controlling transmission timing of the data reproduction processing request packet. A computer program for performing data reproduction control processing based on transmission of a data reproduction processing request to a node connected to a network,
Rule determination for changing the value of the return probability β applied to the process of determining whether or not to execute the process according to the processing request at the receiving node of the data reproduction process request packet in accordance with the data reproduction process request packet transmission sequence A condition setting step;
A packet generation step of generating a data reproduction processing request packet;
A phase control step of controlling a transmission timing of the data reproduction processing request packet,
A computer program comprising:

Images