JP2005070987A - Node to be used in file sharing system, file storage method to be used in file sharing system, and file storing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide new technology to solve the problem that the kinds of files to be shared are reduced by making nodes connected to a network relay a file transfer request, and searching a file whose transfer is requested, and transferring the searched file between nodes to transfer the node to the file request origin. <P>SOLUTION: When the file is transferred, the scale of the demand of the file, the number of nodes which are required to reach the node of the file request origin, the presence probability of its own node, or the number of nodes connected to its own node are examined, and the storage probability of the file is calculated on the basis of the examination result. Then, whether or not the transferred file should be stored is decided on the basis of the storage probability. Thus, it is not necessary to uniquely store a file like the conventional technology. Therefore, it is possible to prevent the reduction of the kinds of files to be shared by effectively using a memory. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a node used in a file sharing system that distributes and stores and uses files at each node connected to the network, a file storage method used in the file sharing system, and a node used in the file sharing system. The present invention relates to a file storage program used for realization.

  There is a client / server type system as a file sharing system via a network. In this system, all files to be shared are stored in the server, and when using the shared file, the client accesses the server. This system has a drawback that a server must be constructed and operated in order to share files, and access to the server is concentrated.

  Another file sharing system via a network is a P2P system. The P2P system copes with the above problem by placing a server in the center. In this system, a file is not stored in a specific server, but is stored in each node constituting the network. That is, the file is distributed and stored in the network.

  P2P includes Hybrid type and Pure type. Hybrid type is a typical system of Napster, and Pure type is a typical system of Gnutella.

  In the hybrid type P2P, a server exists at the center. When a node searches for a specific file, a search request is sent to the server. The server returns to which node the search request has been made, to which node the search target file is stored. The search target file is directly transferred from the node that stores the file to the node that made the search request.

  In Pure type P2P, nodes are connected to each other. When there is a search request for a specific file from a certain node, the search request is transmitted in a bucket relay manner through the network via the connected node. When a search target file is found at a node, the file is directly transferred to the node that issued the search request.

  One example of a P2P file sharing system is Freenet (for example, see Non-Patent Document 1 and Non-Patent Document 2 below).

  Freenet is one of the Pure P2P systems. Freenet is characterized by higher user anonymity and efficient content distribution than the conventional P2P file sharing system by Napster and Gnutella.

  In Freenet, in order to specify a file, Key is assigned to a shared file (Non-patent Document 2: p.4 1.21 to 24). Key has hashed the contents of the content (Non-Patent Document 2: p.5 1.29 to 1.36), and has hashed the structure of the content (Non-Patent Document 2: p.4 1.42 to p. 4). 5 1.4). Each node (terminal) constituting the Freenet holds the Key information of the file held in the information recording device in the own node (Non-patent Document 2: p.3 1.30 to p.3 1.33).

  When searching for a file, the node that performs the search compares the ID recorded in the information recording device of the node with the ID of the search target file. As a result of the comparison, the node that is searching makes a search request to the node that records an ID close to the search target file ID. Thus, the search is performed in a bucket relay system (Non-Patent Document 1: p.21 1.10 to 1.15 / p.22 Figure 5.2, Non-Patent Document 2: p.6 1.19 to 1.24 / p.7 fig.1). .

  When a search target file is found, the file is transferred to the node that performed the search first. At this time, the search target file is stored not only at the node that first performed the search, but also at the relay node that performed the bucket relay type search (Non-patent Document 1: p.21 1.20 to p.22 1.4 / p.22). fig5.2, Non-Patent Document 2: p.6 1.24 to 1.27 / p.7 fig.1).

  At this time, the information recording device may run out of free space, but the most recently saved file in the information recording device is deleted (Non-patent Document 1: p.23 1.10 to 1.25 / p.24 Figure). 5.3, Non-Patent Document 2: p.9 1.19-1.26).

  One characteristic of Freenet is anonymity. This is because the file is stored also in the relay node, and the node holding the file is not the direct transmission node / reception node of the file (Non-Patent Document 2: p.16 1.14 to 1.15).

Another feature of Freenet is efficient content distribution. This is because the file to be searched is duplicated in the network (Non-patent document 1: p.17 1.7 to 1.22, Non-patent document 2: p.7 1.11 to p.8 1.11).
: "A Distributed Decentralized Information Storage and Retrieval System" by Ian Clark, Master's thesis, University of Edinburgh, 1999. : Ian Clark, Oskar Sandberg, Brandon Wiley, Theodore W. Hong "Freenet: A Distributed Anonymous Information Storage and Retrieval System", In Proc. International Workshop on Design Issues in Anonymity and Unobservability, volume 2009 of LNCS, pages 46-66 .Springer-Verlag, 2001.

  A problem of the prior art is that the types of files stored in the network are reduced.

  Files that are in demand, that is, files that are retrieved and downloaded from many nodes, are also cached by many relay nodes compared to files that are not in demand. Since each node has a limited recording capacity to be cached, inevitably less demanding files are deleted from the cache.

  If there is a file on the network that is temporarily in high demand, the file is replicated in large quantities on the network during periods of demand. As a side effect, files that are not in demand from the network are deleted. That is, when considering the entire network, if files that are in temporary demand exist in the network, the types of files existing in the network decrease (Non-Patent Document 1: p.18 1.13 to 1.32, Non-Patent Document 2: p. .9 1.32 to 1.42).

  The present invention has been made in view of such circumstances, and is a file sharing system in which files are distributed and stored and used at each node connected to a network such as P2P, and the number of types of files to be shared is small. The purpose is to provide a new file sharing technology that solves this problem.

  In order to achieve this object, in the node of the present invention, a node connected to a network relays a file transfer request, finds a file with a transfer request, and transfers the found file between the nodes. And a calculation means for calculating the storage probability of the transferred file when used in the file sharing system configured to transfer to the file requesting node, and based on the storage probability calculated by the calculation means. Determining means for deciding whether or not to save the transferred file, and storing the transferred file only when it is decided to save the transferred file according to the decision of the deciding means. Storage means for storing in the means.

  In such a configuration, there may be provided an investigation means for investigating information about the transferred file and an investigation means for investigating the state of the own node. The file storage probability is calculated based on the file information and the node state of the investigation result.

  Further, the storage unit may include an acquisition unit that acquires information used for calculating the storage probability of the file from a storage probability calculation support device connected to the network. In this case, the calculation unit may store the information acquired by the acquisition unit. Will be used to calculate the save probability of the file.

  Each of the processing means described above can be realized by a computer program, and this computer program can be provided by being recorded on a recording medium such as a semiconductor memory or via a network.

  When the file to be transferred is transferred to the file request source node, the node of the present invention configured as described above investigates the magnitude of demand for the transferred file or reaches the file request source node. The number of nodes up to the current node, the existence probability of the own node, the number of nodes connected to the own node, and the storage probability of the transferred file based on the result of the investigation. . At this time, information used for calculating the storage probability of the file is obtained from a storage probability calculation support device connected to the network, and the storage probability of the transferred file may be calculated using the information.

  The file storage probability calculated in this way has a property that it varies depending on the property of the file, and also has a property that it varies for each node, and can be operated from the outside. Has properties.

  Then, when calculating the storage probability, the node of the present invention generates a random value, for example, and determines whether or not to store the transferred file based on the generated random value and the storage probability. Only when the decision is made and it is decided to save the transferred file, the transferred file is stored in the storage means.

  In this way, in the present invention, a node connected to the network relays a file transfer request, finds a file having a transfer request, and transfers the found file between the nodes. Instead of adopting a configuration in which the relay node always stores the transferred file as in the prior art, the file is stored according to the file storage probability. It is determined whether or not to perform, and based on the determination result, the transferred file is saved as necessary.

  In the prior art, a file (content) that is not accessed has been deleted and cannot be obtained. On the other hand, in the present invention, files are deleted less than in the prior art even if there is no access.

  Therefore, according to the present invention, in the file sharing system in which files are distributed and stored and used at each node connected to the network, there is an effect that the types of files existing in the entire network are maintained.

  In the prior art, a node having a specific file is surely related to the file by transmitting / receiving / relaying the file. On the other hand, in the present invention, even a node that does not have a specific file may have transmitted, received, or relayed the file. That is, a specific file is not saved, so that no trace is left, but the file may be received / relayed.

  Therefore, according to the present invention, it is possible to exchange files with more anonymity in a file sharing system in which files are distributed and stored and used in each node connected to the network.

  FIG. 1 illustrates an example embodiment of the present invention. In the figure, 10, 18 and 19 are nodes having the present invention, 16 is a storage probability calculation auxiliary device (device prepared to support storage probability calculation), and 17 is a network.

  The node 10 having the present invention includes a file information investigation device 11, a node state investigation device 12, a storage probability calculation device 13, an information recording device 14, and a communication device 15.

  The node 10 configured as described above is connected to an external network 17 via the communication device 15, and other nodes 18 and 19, a storage probability calculation auxiliary device 16, and the like are connected to the network 17.

  As shown in FIG. 2, the file information investigation device 11 includes a file reception unit 21 that receives a file to be investigated, a hop number investigation unit 22 that investigates the number of hops of a file, and a file demand investigation unit that investigates the demand for a file. And a survey result transmission unit 24 that transmits the surveyed hop count and file demand to the storage probability calculation device 13.

  Here, the file demand is a demand on the network 17 of the search target file. For example, the demand for the file is calculated by counting up the number of times the file has been requested so far. The number of hops is the number of nodes from the requesting node to the local node. For example, when an instruction to transfer the search request up to the number of nodes is given in the search request, each node transfers the search request while reducing the specified number of nodes by one. Therefore, the number of hops is specified from the reduced number of nodes.

  As shown in FIG. 3, the node state investigation device 12 includes an existence probability investigation unit 31 that examines the existence probability of a node, a connection node number investigation unit 32 that investigates the number of nodes connected to the node, And an investigation result transmission unit 33 that transmits the number of connected nodes to the storage probability calculation device 13.

  Here, the existence probability is calculated by the time during which the own node is started up in the past fixed period. The number of connected nodes is the number of other nodes to which the own node is connected via the network 17.

  As shown in FIG. 4, the storage probability calculation device 13 receives a storage probability calculation element reception unit 41 that receives a calculation element for calculating a storage probability from the file information inspection device 11, the node state inspection device 12, and the like. A storage probability calculation unit 42 that calculates the storage probability of the file using the calculated elements, and a storage probability transmission unit 43 that transmits the calculated storage probability to the information recording device 14.

  As shown in FIG. 5, the information recording device 14 includes a storage probability receiving unit 51 that receives the storage probability calculated by the storage probability calculating device 13, and a storage determination unit 52 that determines whether to save the file from the storage probability. A file content recording unit 53 that stores the file when the file is stored, a file receiving unit 54 that receives the file to be stored from the communication device 15, and a file that is stored in the file content recording unit 53 as the communication device 15. And a file transmission unit 55 for transmission to the network.

  A file handled in the network 17 is assigned a unique and unique ID in the network 17. The ID may be obtained by hashing the file name or content, or may be assigned to the file from the outside.

  As shown in FIG. 6, the information recording device 14 of each node is associated with the file ID and information indicating the file data and the location from which the file has been transferred (the file transfer source). The address of the original node).

  In order to effectively use the resources of the information recording device 14, as shown in FIG. 6, the information recording device 14 records the ID of the file and an address indicating from which node the file came from. In addition, a second floor 62 is provided for recording not only the file ID and address but also the data of the file. The capacity of the information recording device 14 and the balance between the first floor portion 61 and the second floor portion 62 are different for each node.

  The communication device 15 is not only a file to be transferred between nodes, but also information on which node the file is stored on and information on what storage probability calculation element is used by other nodes. Communicate with each other.

  As shown in FIG. 7, the storage probability calculation auxiliary device 16 includes a storage probability calculation element input unit 71 that allows the storage probability calculation device 13 to input a calculation element used when calculating the storage probability from the outside, A storage probability calculation element receiving unit 72 for receiving from a specific node data that can be a storage probability calculation element such as what kind of storage probability calculation element is used by a specific node, whether a file is stored, and the input storage A storage probability calculation element transmission unit 73 that transmits a probability calculation element or the like to a specific node via the network 17;

  When searching for a file when configured in this way, the node 10 that performs the search sends the ID of the search target file to the communication device 15 of its own node. Then, the ID of the search target file is compared with the ID recorded in the information recording device 14.

  When the search target file is not found, the node 10 searches the information recording device 14 for a node in which an ID close to the ID of the search target file is recorded, and makes a search request to the node 18, for example.

  For example, when the information recording device 14 records information as shown in FIG. 6 and the search target file is not found, the ID “91a33” recorded in the information recording device 14 is When the ID is closest to the ID of the search target file, for example, the search is requested to the node (for example, the node 18) having the address “127.0.03.1”.

  The node 18 that has received the search request performs a search in the same manner as the node 10. Thus, the search is performed in a bucket relay manner as in the prior art.

  When a search target file is found, the file is transferred to the node that performed the search first. During this transfer, the search target file is stochastically stored not only at the node where the search is first performed but also at the relay node where the search is performed in the bucket relay system.

  In other words, the present invention adopts a configuration in which the search target file is not stored in the relay node that has performed the search in the bucket relay system, but is stored probabilistically.

  FIG. 8 illustrates a processing flow of a node that is executed when a search target file is transferred.

  In the present invention, when a search target file is transferred, the node first receives the transferred search target file as shown in the processing flow of FIG. 8 (procedure indicated by 81 in the figure). ).

  Subsequently, the storage probability calculation element obtained from the file information inspection device 11, the node state inspection device 12, or the storage probability calculation auxiliary device 16 is referred to (procedure indicated by 82 in the figure), and the search transferred based on the storage probability calculation element. The storage probability of the target file is calculated (procedure shown at 83 in the figure). For example, a storage probability such as 35% indicated by a numerical value in the range of 0 to 100 is calculated.

  Subsequently, for example, whether or not to save the transferred search target file based on the generated random number value and the calculated storage probability based on the generated random number value in the range of 0 to 100 Is determined (procedure shown at 84 in the figure). For example, when a storage probability such as 35% is calculated and a random number value that falls within the range of 0 to 35 is generated, it is determined to store the transferred search target file, and 36 to 100 When a random number value that falls within the range is generated, it is determined that the transferred search target file is not saved.

  If it is determined to save the transferred search target file according to this determination, the transferred search target file is stored in the information recording device 14 (procedure indicated by 85 in the figure), and then the next node. (Procedure 87 in the figure).

  On the other hand, if it is determined not to save the transferred search target file according to this determination, the transferred search target file is not stored in the information recording device 14 (procedure indicated by 86 in the figure). Immediately forward to the next node (procedure shown at 87 in the figure).

  In this way, the present invention adopts a configuration in which the search target file is stored probabilistically instead of always storing the search target file at the relay node that has performed the search by the bucket relay method.

  As a result, in a file sharing system in which files are distributed and stored and used at each node connected to the network, the type of files existing in the entire network can be maintained, and more It becomes possible to achieve an effect that file exchange with anonymity is possible.

  According to the present embodiment, the relay node determines whether or not to store the transferred search target file based on a certain storage probability defined in a file-independent form. When this configuration is followed, the file information investigation device 11 / node state investigation device 12 / storage probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

  The operation in this embodiment is as follows.

  The search target file is received by the communication device 15. The storage probability used for determining whether to store the transferred search target file is constant depending on the node. The storage probability calculation device 13 refers to the certain storage probability.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  In the case of following this embodiment, the relay node determines whether or not to store the transferred search target file, based on the number of hops that is the investigation result of the file information investigation apparatus 11. When this configuration is followed, the node state investigation device 12 / preservation probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

  Here, the hop count is the number of nodes from the node that requested the search to the local node. The number of hops is calculated by counting the number of nodes that have actually been relayed until the search target file is found and reach the node from the node that requested the search. Further, when a plurality of nodes are found in a loop or the like during the search, the calculation is performed by obtaining the shortest number of relay nodes.

  The operation in this embodiment is as follows.

The search target file is received by the communication device 15. The file information investigation device 11 investigates the hop number H of the received file. The storage probability calculation device 13 obtains a file storage probability P from the hop count H. For example,
P = a * H + b (a and b are constants. For example, P = −0.1 * H + 1)
Or to have every other relay node save a file,
P = H mod 2 (preservation probability is alternately 0% or 100%)
Ask for it.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  There are several ways of thinking about the relational expression between the number of hops H and the storage probability P, and which relational expression is used is determined depending on which idea is adopted.

  For example, if a search request for the same search target file is often issued again at a node close to the node that requested the search, a relational expression that increases the storage probability as the number of hops decreases is obtained. By using it, the search target file is stored in a node close to the node that requested the search.

  In addition, by storing the search target file in the intermediate node between the node that issued the search request and the node that stored the search target file, the next file request can be quickly performed. When it is desired to be able to search for a search target file, a relational expression is used such that the closer to the intermediate node, the greater the storage probability.

  According to the present embodiment, the relay node determines whether or not to store the transferred search target file based on the file demand that is the investigation result of the file information investigation apparatus 11. When this configuration is followed, the node state investigation device 12 / preservation probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

  Here, the file demand is a demand on the network 17 of the search target file. File demand is set for each file. When a search request is made for a file, the information recording device 14 increases the value to obtain the file demand.

  The operation in this embodiment is as follows.

The search target file is received by the communication device 15. The file information investigation device 11 investigates the file demand D of the received file. The storage probability calculation device 13 obtains a file storage probability P from the file demand D. For example,
P = a * D + b (a and b are constants. For example, P = −0.1 * D + 1)
Ask for it.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  There are several ways of thinking about the relational expression between the file demand D and the storage probability P, and which relational expression is used is determined depending on which way of thinking is adopted.

  For example, when the idea is to prioritize file demand over the number of file types existing in the entire network, a relational expression is used such that the greater the file demand, the greater the storage probability. On the other hand, when the idea is to prioritize the number of file types existing in the entire network over the file demand, a relational expression is used such that the greater the file demand, the smaller the storage probability.

  When following this embodiment, the relay node determines whether or not to store the transferred search target file based on the existence probability of the node that is the investigation result of the node state investigation device 12. When this configuration is followed, the file information investigation device 11 / storage probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

  Here, the existence probability of a node is calculated by the time during which the own node is started up during a past fixed period. The existence probability is calculated by dividing the time of starting the own node by the time after the system is installed. Moreover, you may obtain | require from the starting time in the past one day. Alternatively, it may be obtained from the activation time in a specific condition period (such as every Sunday afternoon).

  The operation in this embodiment is as follows.

The search target file is received by the communication device 15. The node state investigation device 12 investigates the existence probability E of the node. The storage probability calculation device 13 obtains a file storage probability P from the node existence probability E. For example,
P = a * E + b (a and b are constants. For example, P = −E + 1)
Ask for it.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  There are several ways of thinking about the relational expression between the node existence probability E and the storage probability P, and which relational expression is used is determined depending on which idea is adopted.

  For example, in a case where the search target file is preferentially stored in a node that exists on the network for a long period of time, a relational expression in which the storage probability increases as the node existence probability E increases. Will be used. On the other hand, in the case where the search target file is preferentially stored in a node that does not exist on the network for a long period of time, a relational expression that increases the storage probability as the node existence probability E decreases. Will be used.

  In the case of following this embodiment, the relay node determines whether to save the transferred search target file, based on the number of connected nodes that is the investigation result of the node state investigation device 12. When this configuration is followed, the file information investigation device 11 / storage probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

  Here, the number of connected nodes is the number of other nodes to which the own node is connected via the network 17. The number of connected nodes is calculated by counting the number of nodes that have been connected in the past. Alternatively, it may be calculated by counting the number of nodes that are currently connected in the network 17. Alternatively, the calculation may be performed by counting the number of nodes recorded in the information recording device 14.

  The operation in this embodiment is as follows.

The search target file is received by the communication device 15. The node state investigation device 12 investigates the number C of connected nodes of the node. The storage probability calculation device 13 obtains a file storage probability P from the number C of connected nodes. For example,
P = a * C + b (a and b are constants. For example, P = −0.1 * C + 1)
Ask for it.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  There are several ways of thinking about the relational expression between the number C of connected nodes and the storage probability P, and which relational expression is used is determined depending on which idea is adopted.

  For example, in a case where the search target file is actively stored in a node having a large number of connected nodes C, a relational expression that increases the storage probability as the number of connected nodes C increases is used. Will be used. On the other hand, a search target file is frequently transferred to a node having a large number of connected nodes C. From the standpoint of giving priority to the number of file types existing in the entire network, a node having a large number of connected nodes C is used. Is based on the idea of not actively storing the search target file, a relational expression is used such that the storage probability decreases as the number C of connected nodes increases.

  In the case of following this embodiment, the relay node determines whether or not to store the transferred search target file, the number of hops H and the file demand D, which are the investigation results of the file information investigation device 11, and the node state investigation device 12. It is determined by the judgment value J calculated based on the existence probability E and the number of connected nodes C as the investigation result. When this configuration is followed, the storage probability calculation auxiliary device 16 shown in FIG. 1 does not operate.

Here, the judgment value J is, for example,
J = a * H + b * D + c * E + d * C where a, b, c and d are constants
(For example, J = 0.1 * H + 0.2 * D + 0.5 * E-0.05 * C)
Calculate as follows.

  The operation in this embodiment is as follows.

  The search target file is received by the communication device 15. The file information investigation device 11 calculates the file hop count H and the file demand D. The node state investigation device 12 calculates the existence probability E of nodes and the number C of connected nodes.

The storage probability calculation device 13 first calculates a judgment value J, and then obtains a file storage probability P from the judgment value J. For example,
P = a * J + b (a and b are constants. For example, P = −0.1 * J + 1)
Ask for it.

  Whether or not to save the search target file is determined by the information recording device 14 using the storage probability. When storing the search target file, the search target file is recorded by the information recording device 14. Thereafter, the search target file is transferred to another node by the communication device 15.

  In order to assist the storage probability calculation device 13, a storage probability calculation auxiliary device 16 exists outside the node. The storage probability calculation auxiliary device 16 is a device that assists the node when the node calculates the storage probability of the search target file.

  In this embodiment, the storage probability calculation auxiliary device 16 is used to give a storage probability to the node from the outside by the storage probability calculation auxiliary device 16. When this configuration is followed, the file information investigation device 11 / node state investigation device 12 shown in FIG. 1 does not operate.

  The storage probability of each node is basically calculated for each node. However, there is a case where it is desired to give a storage probability to a specific file of each node from the outside. In order to meet such a demand, in this embodiment, the storage probability calculation auxiliary device 16 is used to directly give the storage probability of each node so that the storage probability can be designated.

  The operation in this embodiment is as follows.

  First, the storage probability is input to the storage probability calculation auxiliary device 16. The inputted storage probability is transmitted to the communication device 15 of the node 10 via the network 17. The storage probability transmitted in this way is transmitted from the communication device 15 to the storage probability calculation device 13. When each node is given a storage probability from the storage probability calculation auxiliary device 16, the storage probability is preferentially calculated.

  According to the present embodiment, the operator can specify that the search target file is stored on the network and the search target file is not stored on the network.

  Also in this embodiment, the storage probability calculation auxiliary device 16 is used. In this embodiment, the storage probability calculation auxiliary device 16 does not give a storage probability to the node from the outside, but gives a calculation element when calculating the storage probability.

  According to the present embodiment, the storage probability calculation device 13 uses the storage probability calculation elements such as the number of hops, the file demand, the existence probability, and the number of connected nodes, and the storage probability calculation elements given from the storage probability calculation auxiliary device 16 as variables. Use the function to calculate the storage probability. Alternatively, the storage probability is calculated by correcting the storage probability calculated from the storage probability calculation elements such as the number of hops, the file demand, the existence probability, and the number of connected nodes with the storage probability calculation auxiliary device 16. You may do it.

  The operation in this embodiment is as follows.

  First, a storage probability calculation element to be given to the node is input to the storage probability calculation auxiliary device 16. The inputted storage probability calculation element is transmitted to the communication device 15 of the node 10 via the network 17. The storage probability calculation element transmitted in this way is transmitted from the communication device 15 to the storage probability calculation device 13. When each node is given a storage probability calculation element from the storage probability calculation auxiliary device 16, the storage probability is calculated by giving priority to (considering) these storage probability calculation elements.

  According to this embodiment, the storage probability calculated by each node can be controlled by the storage probability calculation element given from the outside.

  For example, as an example, the storage probability calculation auxiliary device 16 notifies the node that is trying to calculate the storage probability of the storage probability calculation element used by a node (including a node other than the relay node) located in the vicinity thereof. Thus, it is possible to control so that the storage probability in a form consistent with the storage probability calculated by the nodes located in the vicinity is calculated.

  In this embodiment, the storage probability calculation assisting device 16 notifies the storage probability calculation element to the node for which the storage probability is to be calculated. Information that is helpful in calculating the storage probability may be notified, such as notifying what file is stored (including nodes other than the node).

  For example, if the storage probability calculation auxiliary device 16 notifies what kind of file is stored in a nearby node, the storage probability calculation device 13 of the node for which the storage probability is to be calculated is, for example, When a search target file is already stored in a node located in the vicinity, the search target file is controlled not to be stored in its own node by reducing the storage probability of the search target file. Processing can also be performed. Accordingly, each node can calculate an appropriate storage probability by calculating the storage probability of the search target file in consideration of information about nodes other than the node itself.

1 is an example embodiment of the present invention. It is an example of the apparatus structure of a file information investigation apparatus. It is an example of the apparatus structure of a node state investigation apparatus. It is an example of the apparatus structure of a preservation | save probability calculation apparatus. It is an example of the apparatus structure of an information recording device. It is explanatory drawing of the information which an information recording device preserve | saves. It is an example of the apparatus structure of a preservation | save probability calculation auxiliary | assistance apparatus. It is an example of the processing flow which this invention performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Node 11 File information investigation apparatus 12 Node state investigation apparatus 13 Storage probability calculation apparatus 14 Information recording apparatus 15 Communication apparatus 16 Storage probability calculation assistance apparatus 17 Network 18 Node 19 node

Claims (10)

The node connected to the network relays the file transfer request, finds the file with the transfer request, and transfers the found file between the nodes to transfer it to the file requesting node. A node used in a file sharing system
Means for deciding whether to save the transferred file;
Means for storing the transferred file in the storage means only when it is determined to save the transferred file in accordance with the determination.
Node used in the featured file sharing system. The node connected to the network relays the file transfer request, finds the file with the transfer request, and transfers the found file between the nodes to transfer it to the file requesting node. A node used in a file sharing system
Means for calculating the storage probability of the transferred file;
Means for determining whether to save the transferred file based on the storage probability;
Means for storing the transferred file in the storage means only when it is determined to save the transferred file in accordance with the determination.
Node used in the featured file sharing system. In the node used in the file sharing system according to claim 2,
The means for determining generates a random value and determines whether to save the transferred file based on the generated random value and the storage probability.
Node used in the featured file sharing system. In the node used in the file sharing system according to claim 2 or 3,
It has a means to investigate information about transferred files,
The calculating means calculates the file storage probability based on the file information of the survey result.
Node used in the featured file sharing system. In the node used in the file sharing system according to claim 2 or 3,
It has a means to investigate the status of its own node,
The calculating means calculates the file storage probability based on the node state of the investigation result.
Node used in the featured file sharing system. In the node used in the file sharing system according to claim 2 or 3,
A means of investigating information about transferred files;
And a means for investigating the state of the own node,
The calculating means calculates the file storage probability based on the file information of the investigation result and the node state of the investigation result.
Node used in the featured file sharing system. In the node used in the file sharing system according to any one of claims 2 to 6,
From the storage probability calculation support device connected to the network, comprising means for obtaining information used for calculating the storage probability of the file,
The calculating means calculates the file storage probability using the acquired information.
Node used in the featured file sharing system. The node connected to the network relays the file transfer request, finds the file with the transfer request, and transfers the found file between the nodes to transfer it to the file requesting node. A file storage method used in a file sharing system,
Only when the node receiving the file to be transferred to the file requesting node determines whether to save the transferred file and, according to the determination, determines to save the transferred file. Saving the transferred file to the storage means
A file storage method used in a featured file sharing system. The node connected to the network relays the file transfer request, finds the file with the transfer request, and transfers the found file between the nodes to transfer it to the file requesting node. A file storage method used in a file sharing system,
The node that receives the file to be transferred to the file requesting node calculates the save probability of the transferred file, and determines whether to save the transferred file based on the calculated save probability. Storing the transferred file in the storage means only when it is determined to save the transferred file according to the determination,
A file storage method used in a featured file sharing system.   A file storage program for causing a computer to execute processing used to realize a node used in the file sharing system according to claim 1.

Images