JP5119844B2 - File transfer system, file transfer method, file transfer program, and index server - Google Patents

Description

  The present invention relates to a file transfer system, a file transfer method, a file transfer program, and an index server. For example, the present invention is suitable for application when transferring the same file to many nodes.

  For example, in a content distribution system, the same content (file) is distributed to many terminals. In such a system, since the capabilities and specifications of the terminals vary, content is distributed in an environment where mutual communication is possible using a general-purpose network protocol centering on IP (Internet Protocol). For example, a large-capacity content exceeding 10 MB is difficult to transfer, and the resources on the transmission source device (server) side are concentrated on the transfer process during the transfer. Are likely to become congested or the processing of the transmission source device cannot catch up and the service becomes impossible.

  Therefore, it is desired to reduce the load on the transmission source device and avoid congestion. In addition, it is desired to distribute a file after dividing it into an appropriate size so that it can be repaired even if the transfer is interrupted.

  In order to meet such a demand, it is considered to use a P2P type file exchange technology.

  In a client / server type system, since the file exists in the server, it is not necessary to search for the “location” of the file.

On the other hand, in the P2P system, the file is held by each peer, not the file is exist collectively in a fixed location. Therefore, a node that requests a file needs “data search” to obtain the location information of the file by some method before starting to acquire the file. P2P applications have various contrivances in the search method, and can be classified into (1) hybrid P2P, (2) pure P2P, (3) super node type hybrid P2P, etc. (see Non-Patent Document 1). ).

  “Hybrid P2P” is a hybrid of a client / server type and a P2P type. File location information is handled by a fixed server. Each node knows the location of the file by making an inquiry to a server that it knows in advance. A server that searches for a file is called an “index server”. The index server handles only the file location information, and the file entity is different from the client / server type in that peers communicate with each other.

Patent Document 1 describes a technology in which such a hybrid P2P is applied to a content distribution system. The system described in Patent Literature 1 includes a large number of dynamic peers that mutually distribute content and a center server that manages the distribution of content. Each dynamic peer notifies the center server of its operation status, and the center server calculates and registers the load of each dynamic peer based on the operation status. Each dynamic peer acquires a content list from the center server, searches for other dynamic peers that have acquired the desired content, and downloads content from the dynamic peer with the lowest load.
http: // japan. internet. com / column / webtech / 20050511/8. html JP 2006-72432 A

  In the P2P type system described in Patent Document 1, an efficient topology is configured between peers, and file concentration between peers can reduce the problem of load concentration on a server (or static peer).

  However, in the technology described in Patent Document 1, content is replicated and distributed in each dynamic peer, but since the dynamic peer with the smallest load at that time is determined as the distribution source, the center server It is necessary to keep track of the load status of each dynamic peer, and to recognize the dynamic peer holding each content while the load is constantly changing. Is big. Also, dynamic peers and static peers must frequently report the load to the center server.

  Even if the P2P type is applied, when there is one node (server or peer) holding a file, access must be concentrated on the node having the file.

  In addition, P2P type systems have problems in guaranteeing the legitimacy of files, such as fake forgery with a file name different from the contents, and the distribution of malicious files mixed with viruses.・ The problem that the distribution of information that is not desirable for distribution, such as national secrets and corporate secrets, cannot be stopped once the file has spread, has become apparent from numerous incidents such as recent information leaks.

  Furthermore, with the conventional file transfer method, there may be a problem that it is not easy to find the transfer source again once the connection with the transfer source is broken once the target file is extremely few in the network. is there. For example, in the case of the technology described in Patent Document 1, when the dynamic peer that has received the content transfer loses the information of the transfer source due to the disconnection abnormality, information on which transfer source was transferred to the center server Therefore, it is not easy to find the transfer source again, and there is a problem that it is necessary to start again from the determination of the transfer source.

  The present invention has been made in view of the above points, and can solve the problems of the infrastructure system such as network congestion and load concentration on the server. Further, the present invention guarantees validity, prevents undesired information distribution, and efficiency of information distribution. An object of the present invention is to provide a file transfer system, a file transfer method, a file transfer program, and an index server that can achieve the above.

In the first aspect of the present invention, a node that desires a file inquires a node serving as a transfer source node holding the file to the index server, receives a reply of the transfer source node determined by the index server, and receives a response from the transfer source node. In a file transfer system that receives file transfers, the index server manages topology information in a tree structure in which transfer sources and transfer destination nodes in past transfers are connected by links having directions, for each file. Parameters for determining a transfer source node other than the topology structure , and for each node that can be a transfer source node, a parameter management unit for determination that manages at least a region to which the node belongs and an available bandwidth; and a forwarding node determining means for determining the forwarding node, the upper Symbol transfers Node determining means, from the candidates of the source node, bandwidth usage by a node belonging to the neighborhood local node wants the file is large node, and determines on the forwarding node.

In the second aspect of the present invention, a node that desires a file inquires a node serving as a transfer source node holding the file to the index server, receives a reply of the transfer source node determined by the index server, and receives a response from the transfer source node. In an index server in a file transfer system that receives a file transfer, (1) topology information for managing, for each file , topology information of a tree structure in which a transfer source and a transfer destination node in a past transfer are connected by a link having a direction. (2) Parameters for determining a transfer source node other than the topology structure, and for each node that can be a transfer source node, a determination parameter for managing at least a region to which the node belongs and an available bandwidth a management unit, (3) the forwarding node determining means for determining the forwarding node Comprising a, (4) above Symbol forwarding node determining means determines from the candidates of the source node, a large node bandwidth usage is a node belonging to the neighborhood local node wants the file to the forwarding node characterized in that it.

In the third aspect of the present invention, a node that desires a file inquires a node serving as a transfer source node holding the file to the index server, receives a reply of the transfer source node determined by the index server, and receives a response from the transfer source node. In the file transfer method for receiving a file transfer, (1) the index server includes topology information management means, determination parameter management means, and transfer source node determination means, and (2) the topology information management means includes past transfer information. connecting the transfer source and the transfer destination node links having direction in the topology information of the tree structure, and managed for each file, (3) the determined parameter managing means, the forwarding node outside topology structure surface For each node that can be a transfer source node. (4) The transfer source node determination means selects a node that belongs to the vicinity of the node that wants the file and has a large use band from among the transfer source node candidates. The transfer source node is determined .

According to a fourth aspect of the present invention, a node that wants a file inquires a node serving as a transfer source node holding the file to the index server, receives a reply of the transfer source node determined by the index server, and receives a response from the transfer source node. In a file transfer program mounted on an index server of a file transfer system that receives a file transfer, the computer mounted on the index server is connected to (1) a link having a direction between a transfer source and a transfer destination node in the past transfer. Topology information management means for managing the topology information of the connected tree structure for each file, and (2) parameters for determining a transfer source node other than the topology structure surface, and for each node that can be a transfer source node Decision to manage at least the region to which the node belongs and the available bandwidth And use parameter management unit, (3) transfer from the source node of the candidate, a large node bandwidth usage is a node belonging to the neighborhood local node wants the file, the source node determining means for determining to the forwarding node It is made to function as.

  According to the present invention, it is possible to solve infrastructure problems such as network congestion and load concentration on a server, and further, it is possible to achieve guarantee of legitimacy, prevention of undesirable information distribution, and efficiency of information distribution.

(A) First Embodiment Hereinafter, a first embodiment of a file transfer system, a file transfer method, a file transfer program, and an index server according to the present invention will be described in detail with reference to the drawings. Here, the file transfer system according to the first embodiment belongs to the hybrid P2P type in terms of the classification of the P2P type file exchange technology.

(A-1) Configuration of First Embodiment FIG. 2 shows a network connection configuration of components of the file transfer system 10 according to the first embodiment.

  In FIG. 2, the file transfer system 10 according to the first embodiment includes one or a plurality of index servers 12 that grasp all file search information, and a plurality of files that can be transfer sources and transfer destinations of files used by end users. Client terminals (hereinafter abbreviated as terminals) 13 are connected to the Internet 11.

  Each index server 12 can use a general-purpose server as hardware, and each terminal 13 can use a general-purpose terminal as hardware. As a file transfer program according to the first embodiment, each hardware has a function. The software of the terminal 13 is mounted as a P2P application, for example. The index server 12 may be configured by a single index server, and may be configured by distributing a plurality of servers for the purpose of load distribution and redundancy for ensuring reliability (FIG. 2 shows a plurality of servers). Is shown in a distributed deployment). In the case of distributed processing with a plurality of units, it is desirable to synchronize various databases, which will be described later, in real time. For this reason, a single database is shared by a plurality of servers, or a plurality of databases have a redundant configuration. A method may also be applied.

  FIG. 1 is a block diagram showing a logical functional configuration of the index server 12 and a logical functional configuration of the terminal 13.

  As shown in FIG. 1A, the index server 12 includes a control unit 21, a terminal selection unit 22, an information correction unit 23, a log recording / storage unit 24, an input / output unit 25, a terminal database unit 26, and a file database unit 27. Have Note that the communication unit is omitted in FIG.

  The terminal database unit 26 and the file database unit 27 may be degenerated in the index server 12 or may be separate devices. In the case of degeneration, the input / output unit 25 has an in-device interface function such as PCI, ATAPI, SCSI, and USB, and in the case of another device, the input / output unit 25 has a network interface function.

  The control unit 21 includes a CPU and a program executed by the CPU, and controls processing of other functional units via a bus.

  The terminal selection unit 22 selects a connection destination terminal (transfer source terminal) that is optimal for the terminal 13 that has transmitted the transfer source search request, and notifies the terminal 13 that has transmitted the transfer source search request as the designated connection destination. The terminal selection unit 22 is based on the file information related to the transfer source search request, information on the terminal 13 that is obtained by searching the file database unit 27, and terminal information that is obtained by searching the terminal database unit 26 A connection destination terminal that is optimal for the terminal 13 that has transmitted the transfer source search request is selected using (use band, affiliation area information, maximum permissible transfer number per terminal, and current terminal transfer number).

  In the case of the first embodiment, the terminal selection unit 22 basically selects a connection destination terminal (transfer source terminal) according to a method suitable for calling a bucket relay method or a circulation method (hereinafter referred to as a circulation method). (Decision) is made. When the nearest terminal 13 on a straight line by the circulation method cannot become a connection destination terminal in view of connection destination restriction conditions such as a use band, affiliation area information, a maximum allowable transfer number per terminal, and a transfer number of the current terminal. Then, another terminal 13 on the straight line is selected as a candidate, and another terminal 13 that satisfies the connection destination restriction condition is selected as the connection destination terminal (transfer source terminal). Therefore, as described above, the file flow may branch on the upstream side based on the circulation method, and has a tree structure in terms of topology.

  The terminal database unit 26 manages information (terminal information) of each terminal 13 accommodated in the file transfer system 10. As shown in FIG. 3, the terminal information (one record) includes a user ID / password 26a, a terminal address 26b, a use band 26c, a maximum allowable transfer number 26d per file, a maximum allowable transfer number 26e per terminal, a current The number of terminal transfers 26f, affiliation area information 26g, possessed file 26h, and the like.

  The user ID / password 26a is used for user authentication.

  The terminal address 26b is used, for example, to select the one with the smaller number of hops obtained by using ICMP traceroute or the like as the priority connection destination. That is, when there are connection destination candidates at the same level, the connection destination is determined in consideration of the number of hops.

  The use band 26c is the upper limit of the band that the terminal 13 can transfer as a connection destination terminal (transfer source terminal). That is, regardless of the type of file, the total bandwidth transferred as a transfer source cannot exceed this usage bandwidth. The use band 26c is used, for example, to select the terminal 13 having a larger value as the priority connection destination.

  The maximum permissible transfer number 26d per file is the upper limit number per file that can be transferred as a system, in other words, the maximum number of links (branches) that can be transferred. That is, the number of files transferred as a transfer source cannot exceed the maximum allowable transfer number per file. When there are a plurality of similar connection destination candidates, priority can be given to the one having a larger margin for the maximum permissible transfer number per terminal 26e described later. If the maximum permissible transfer number per file 26d is not reached, it is possible to simultaneously receive different portions of information from a plurality of transfer sources.

  The maximum permissible transfer number 26e per terminal is an upper limit of transfer that the terminal 13 can transmit as a transfer source regardless of the file type.

  The current transfer number 26f of the terminal is the total number of files transferred by the terminal 13 as a connection destination terminal (transfer source terminal) up to that point (the total transfer number regardless of the file type) or the transfer number for each file. It is.

  The affiliation area information 26g is information on the area to which the terminal 13 belongs. For example, the terminal selection unit 22 internally stores information between regions that allow file transfer (for example, the same region or a nearby region). The belonging region information 26g is used to select the terminal 13 having the same or nearby belonging region information as a priority connection destination. The affiliation area information 26g may be used to determine whether or not the area is between areas where file transfer is permitted.

  The owned file 26h is the type of file (file name) maintained by the terminal 13 and the number of files transferred.

  In the file database unit 27, topology information 27B related to the file transfer path is stored for each file (for each file type) in addition to the information of the upload list 27A provided from the terminal 13. In addition, the file database unit 27 is also provided with a determination non-reception list 27C.

  Since the file flow is based on the circulation method and may be branched on the upstream side, the topology information 27B has a linear structure or a tree structure branched downward.

  The upload list 27A is a list of files that can be transmitted by each terminal. As shown in FIG. 4, for each file, the file name 27Aa, file size 27Ab, hash information 27Ac, file type 27Ad, search keyword 27Ae, sampling rate It includes information such as related parameters such as 27Af, used codec 27Ag, and resolution 27Ah.

  Each file can be uniquely identified by a set of four parameters: file name 27Aa, file size 27Ab, hash information 27Ac, and file type 27Ad.

  The search keyword 27Ae allows the user to instruct the terminal 13 to search for a file using the search keyword 27Ae. In order to identify the presence of a file including the search keyword 27Ae. Used.

  The file type 27Ad, the sampling rate 27Af, the use codec 27Ag, and the resolution 27Ah are used for identifying whether or not the terminal 13 (user) can prepare a reproducible environment for audio and moving image files. For example, when a reproducible environment is not prepared, it is possible to improve the usability of the user by adding a mechanism for notifying the terminal 13 after the transfer is completed.

  In the determination unreceived list 27C, when a terminal transfers a file as a transfer source, the transfer destination terminal determines the validity of the received file and notifies the index server 12 of the determination. If not, the transfer source terminal is listed in this list 27C.

  The log recording / storage unit 24 accumulates information related to various processes such as transfer and information transmission / reception of the index server 12 and failure events.

  The information correction unit 23 performs calculations and information correction for updating information in the terminal database unit 26 and the file database unit 27.

  On the other hand, as shown in FIG. 1B, each terminal 13 includes a control unit 31, a setting storage unit 32, a log recording / storage unit 33, a file storage unit 34, an input / output unit 35, a partial cache storage unit 36, and a connection. A state monitoring unit 37 is included. Note that the communication unit is omitted in FIG.

  The control unit 31 includes a CPU and a program executed by the CPU, and controls processing of other functional units via a bus.

  The setting storage unit 32 stores settings such as a user ID / password, a terminal address, a use band, and the like for the terminal 13.

The log recording / storage unit 33 accumulates information related to various processes such as transfer and information transmission / reception of the terminal 13 and failure events.

  The file storage unit 34 corresponds to a hard disk or other large-capacity storage device, and holds file information. The file information includes, for example, a file name, file size, hash information, file type, search keyword, related parameters such as sampling rate, codec used, resolution, and the like.

  The input / output unit 35 corresponds to an input unit such as a keyboard and a mouse and an output unit such as a display and a printer. The input / output unit 35 receives a file transfer request from a user, and displays and prints out the transferred file.

  The partial cache storage unit 36 temporarily stores a file being transferred or a file being transferred. When the terminal 13 operates as a transfer source terminal, the corresponding file stored in the file storage unit 34 is moved to the partial cache storage unit 36. When the terminal 13 operates as a transfer destination terminal, the file being received is stored in the partial cache storage unit 36, and the file is moved from the partial cache storage unit 36 to the file storage unit 34 at the end of transfer.

  The connection state monitoring unit 37 relates to the connection state with other terminals 13 and the index server 12 via the Internet 11 and performs retransmission control or the like as necessary.

(A-2) Operation of the First Embodiment Next, various operations of the file transfer system 10 according to the first embodiment will be described. As described above, the index server 12 may operate by distributing the load on a plurality of servers. However, in the following description, the index server 12 will be described as a single unit for the sake of simplicity.

  First, the operation of the index server 12 when the service is started and when the service is stopped will be described.

  The index server 12 in an initial state (service standby state) performs a start-up process in response to a start-up operation, confirms the normality of the start-up process, and enters the service operating state when it is determined to be normal. When it is determined that the normality confirmation result of the activation process is abnormal, the index server 12 maintains the initial state.

  Further, the index server 12 in the operation state performs a service stop process according to the stop operation, confirms the normality of the stop process, and enters an initial state when it is determined to be normal. When it is determined that the normality check result of the stop process is abnormal, the operation state is maintained.

  Next, the operation when the terminal 13 is connected to the system will be described with reference to the flowchart of FIG. The operation of the index server 12 corresponding to the operation when the terminal 13 is connected to the system will also be described.

  When the index server 12 is in the operation state, the terminal 13 makes an authentication request with terminal information to the predetermined index server 12 via the Internet 11 from the system disconnected state (step 101) (step 102). ). Prior to this request, the terminal 13 takes in the user ID and password from the user, and in addition to these, the address of the terminal, the bandwidth used, the maximum allowable transfer number per file to be transferred, and the maximum allowable transfer per terminal. The number and affiliation area information is included in the terminal information and an authentication request is made.

  Here, the index server 12 that has received the authentication request with the terminal information authenticates with the ID and password of the user of the terminal 13 that is the authentication request source, returns the authentication result to the terminal 13, and the terminal 13 The authentication result is confirmed (step 103). If the authentication fails, the index server 12 rejects the connection of the terminal 13, and the terminal 13 maintains the disconnected state based on the authentication result reply. If the authentication is successful, the terminal 13 transmits upload list information to the index server 12 (step 104).

  When the index server 12 receives the upload list information from the terminal 13, the index server 12 checks the file forgery / defective rate based on the upload list information and notifies the terminal 13 of the result (step 105). File creation / defective rate is the total number of files listed in the upload list information as the denominator, and the number of abnormal files such as forgery or defects among the files listed in the upload list information is taken as the numerator. Ratio. When the forgery / defective rate is equal to or greater than a predetermined threshold, the index server 13 rejects the connection of the terminal 13 and the terminal 13 maintains the disconnected state. If the forgery / defective rate is less than the threshold, the terminal 13 shifts to a connected state (step 106).

  In the case of the first embodiment, at the time of file transfer, the terminals 13 that exchange files take a tree-type topology, and the file flow is a circulation type. The transfer destination (downstream side) terminal cannot receive information on a file that exceeds the capacity of the partial cache storage unit 36 of the transfer source (upstream side) terminal.

  Next, the operation when the terminal 13 receives a desired file transfer will be described in detail with reference to the flowcharts of FIGS. FIG. 6 shows the operation on the terminal 13 side, and FIG. 7 shows the operation on the index server side.

  When the terminal 13 is connected to the system (step 151) and the user performs a request for receiving a desired file (for example, the file is specified by a search keyword), the file is transferred. The index server 12 is inquired of another terminal as a connection destination for receiving the message (step 152).

  When the index server 12 receives an inquiry from the terminal 13 in the operational state (step 201) (step 202), the index server 12 checks the presence / absence of a desired file and the presence / absence of an available connection destination terminal (step 203). If the desired file does not exist, a message to that effect is returned to the terminal 13 (step 210). If the desired file exists but there is no available connection destination terminal, a standby instruction is issued. Is returned to the terminal 13 as a response to the inquiry (step 209), and when the desired file exists and there is an available connection destination terminal, the designated connection destination is determined (one candidate is selected). In this case, the designated connection destination information is automatically returned to the terminal 13 as a response to the inquiry (step 204).

  The index server 12 appropriately updates the terminal database unit 26 and the file database unit 27 and stores a log in the log recording / storage unit 24 when replying to the inquiry. For example, the inquiry terminal is added as “receiving terminal” to the topology information of the file database unit 27.

  The case where the desired file exists but there is no usable connection destination terminal means that, for example, a terminal that can be a connection destination terminal is currently transferring and has no parallel transfer capability, or the parallel transfer capability is Even in such a case, a case where a desired file cannot be transferred in the unused bandwidth of the usage bandwidth is applicable.

  The index server 12 that has received the inquiry from the terminal 13 determines the designated connection destination as follows, for example.

  First, it is confirmed whether a holding terminal can be a connection destination. When there is one holding terminal that can be a connection destination, the holding terminal is determined as a connection destination terminal. In the first embodiment, the “holding terminal” refers to a terminal (terminal serving as a top node in a tree structure) when the file is introduced into the file transfer system 10. In the second embodiment to be described later, the “holding terminal” means a terminal that stores the entire file, and not only the terminal when the file is introduced into the file transfer system 10 but also the transfer. The terminal that stores the entire file because it received the message is also included in the concept of the holding terminal.

  When a plurality of holding terminals can be connection destinations, a terminal in a vicinity area (concept including the same area) of the inquiry terminal and a terminal having a large use band is given priority. In the case where there are a plurality of holding terminals having a high priority level in this aspect, it is determined from the values in the various parameters described above. The priority relationship between the various parameters may be arbitrarily determined, and the stage of each parameter may be converted into a numerical value, a weighted addition value of those numerical values may be calculated as an evaluation value, and a higher evaluation value may be selected. In other words, regarding the priority determination algorithm, it is desirable that weighting setting by parameters is possible.

  If all holding terminals have no room to accept new connection requests for transfer to other terminals, or if the maximum allowable transfer number per terminal has been reached and it can no longer be the transfer source, it can become the connection destination. If not, the connection destination is determined from non-holding terminals other than the holding terminal that are not receiving the file. The connection destination selection algorithm when there are a plurality of such terminals is the same as that of the holding terminal, and one connection destination terminal is determined.

  If there is no terminal that is not receiving a file among the non-holding terminals, a connection destination is selected from the receiving terminals. In the selection algorithm of the connection destination from the receiving terminal, priority is given to the terminal of the upper layer (upstream side close to the top node), and the selection algorithm in the holding terminal is applied mutatis mutandis when selecting from a plurality of terminals in the same layer. . One feature of this embodiment is that the receiving terminal is also the transfer source terminal. This is because the terminal being received is given priority to the terminal of the upper layer (upstream side close to the top node), and therefore the received file portion is larger than the terminal of the lower layer.

  When determining the connection destination, the holding terminal or the non-holding terminal that has reached the allocation of the maximum permissible transfer number per file and the maximum permissible transfer number per terminal is excluded from the selection target as the connection destination.

  With the connection destination determination method as described above, when a large number of transfer source search requests are concentrated in a certain period, a transfer path topology having a tree structure by the receiving terminal is formed.

  The terminal 13 replies from the inquired index server 12 that the desired file does not exist or that the desired file exists but there is no usable connection destination terminal (step 153). Return to the state.

  On the other hand, when the terminal 13 returns the designation of the connection destination terminal from the inquired index server 12 (step 154), the terminal 13 requests the designated connection destination terminal to transfer the corresponding file (step 154). 155) Receives the transferred file (step 156) and waits for the completion of the file transfer (step 157).

  If the connection with the connection destination terminal (transfer source terminal) is disconnected before the file transfer is completed (if the file transfer has not been completed), the terminal 13 processes from the inquiry to the index server 12 Try again. Although not shown in FIG. 7, the index server 12 recognizes that the same terminal 13 has inquired again from past logs and the like, and determines the connection destination terminal in consideration of the past determination.

  When the file transfer is completed, the terminal 13 transmits a transfer completion notification (reception completion notification) to the index server 12 (step 158).

  When the index server 12 receives the transfer completion notification (step 205), the index server 12 performs processing upon reception (step 206). For example, the number of transfer of the transmitting terminal is reduced by one, and in the topology information of the file database unit 27, the receiving terminal is rewritten from “receiving terminal” to “non-holding terminal” holding the file, The hierarchical information of the receiving terminal is rewritten on the basis of the corresponding receiving terminal. Further, for example, information of a connection destination terminal (transfer source terminal) is added to a terminal list that has not received a validity determination result.

  The terminal 13 determines the legitimacy of whether the transferred file is forged (e.g., determined from hash information) or a defective file (e.g., a file infected with a virus). The determination result is transmitted to the index server 12 (step 159), and the connection state (standby state) is restored. Various existing methods can be applied as a method of determining validity.

  Upon receiving the validity determination result, the index server 12 performs a process at the time of reception (step 208). For example, the forgery / defective rate information is recalculated, or the connection destination (transfer source) terminal is deleted from the determination non-reception list. Further, when the determination result indicates a forged or defective file, a process of updating information so that the current connection destination (transfer source) terminal cannot be determined as the connection destination terminal in the future is performed.

  A specific example of determining the connection destination terminal (transfer source terminal) will be described with reference to FIG. In FIG. 8, the terminal 13A is a holding terminal, the terminals 13B and 13C receive file transfer from the holding terminal 13A, the terminal 13D receives file transfer from the terminal 13B, and such topology information is stored in the index server 12. This is the case. In such a case, it is assumed that the terminal 13E makes an inquiry to the index server 12.

  If the only holding terminal 13A can be the connection destination terminal, the holding terminal 13A is determined as the connection destination terminal.

  For example, if the maximum allowable transfer number per file of the holding terminal 13A is “2” and the holding terminal 13A can no longer be a connection destination terminal, the index server 12 connects from the terminals 13B to 13D. Determine the destination terminal. First, it is confirmed whether or not any of the terminals 13B to 13D has completed file reception. If only the terminal 13B has received the file, this terminal 13B is determined as the connection destination terminal. If the files having been received are the terminals 13B and 13C (the terminal 13D is receiving), the terminals 13B and 13C are first compared with the consistency of the area with the terminal 13E and the amount of bandwidth used. One terminal 13B or 13C is determined as a connection destination terminal. If the area consistency and the usage bandwidth are at the same level, a terminal to be prioritized is determined from other parameters.

  Further, for example, if the holding terminal 13A and the terminal 13B can no longer be the connection destination terminals, the index server 12 determines the connection destination terminal from the terminals 13C and 13D. When both the terminals 13C and 13D have completed the file reception, the connection destination is determined in the same manner as when the terminals that have completed the file reception are the terminals 13B and 13C, regardless of the difference in the hierarchy. Determine the terminal. When the terminals 13C and 13D are both receiving terminals, the terminal 13C is determined to be the connection destination terminal because the terminal 13C is a higher layer.

For example, if it is determined the terminal 13C as the connection destination terminal of the terminal 13E is the topology information, links connecting terminals 13E and 13C are added, also, updating of the information of the terminals 13E and 13C in the terminal database unit 26 Etc. are executed.

  Although illustration of the flowchart is omitted, in a certain terminal 13, when a certain file is deleted from the file storage unit 34, the fact is notified from the terminal 13 to the index server 12, and the index server 12 As a trigger, information indicating that the corresponding terminal 13 holds the file is deleted from the file database unit 27.

  Next, the operation of the terminal 13 when the system is disconnected will be described. When a disconnection operation is performed while the terminal 13 is in a connected state (standby state), a disconnection request is made to a predetermined index server 12 via the Internet 11. When receiving the disconnection request, the index server 12 performs system disconnection processing and transmits a disconnection response to the terminal 13. When receiving the disconnection response, the terminal 13 enters a disconnected state. Further, if a failure occurs in the network or system when the terminal 13 is connected or disconnected, the terminal 13 is forcibly disconnected.

(A-3) Effects of the First Embodiment According to the first embodiment, since the file transfer system belongs to the hybrid P2P type, it is possible to solve infrastructure problems such as network congestion and load concentration on servers. That is, since the file holding terminals are distributed, congestion due to concentration of access can be avoided.

  In addition, since information related to file transfer can be managed on the index server side, it is easy to prevent the circulation of harmful files containing illegal files (for example, copyright infringing files) or viruses.

  Furthermore, since file distribution is based on a tree-like topology, distribution efficiency can be achieved, and since downloaded users can be recorded, it is easy to manage file (information) distribution.

  Furthermore, unlike existing P2P applications, the topology configuration between terminals is not unique for all file searches and downloads, and a different configuration for each transfer file occurs autonomously. Can be used.

  Even if the file transfer is interrupted in the middle, the connection destination can be easily searched based on the log information of the index server or the like.

(B) Second Embodiment Next, a file transfer system, a file transfer method, a file transfer program, and an index server according to a second embodiment of the present invention will be described with reference to the drawings.

  In the first embodiment described above, a terminal that can be a file transfer source is a terminal other than the holding terminal and the holding terminal when the file is introduced into the file transfer system 10 and holds the entire file. The terminals are classified into three categories, that is, the terminal that is receiving the file and the terminal that is receiving the file. Differences in these categories also affect the determination of the connection destination terminal.

In the second embodiment, the entire file is stored because it is a terminal when the file is introduced into the file transfer system 10, or the entire file is stored because the file transfer is received. Regardless of whether the terminal stores the entire file is the holding terminal. Therefore, in the second embodiment, there are only two classifications of the holding terminal and the receiving terminal, and the difference between the two classifications is reflected in the determination of the connection destination terminal. In between the holding terminal and the reception in the terminal, if less than the file per maximum allowable number of transfers is set by the index server 12 (the number of transfer chain) is to prioritize the holding terminal.

  The index server 12 may execute the process of rewriting “receiving terminal” to “holding terminal” in the process of step 206 in FIG. 7 described above when receiving the file transfer completion notification. Others are the same as in the first embodiment.

  Rewriting “receiving terminal” to “holding terminal” means that if there is a terminal receiving file transfer with the receiving terminal as the transfer source terminal as shown in FIG. 9, the topology related to that terminal is also changed. It will be the same as

  According to the second embodiment, the same effect as that of the first embodiment can be obtained. In addition to this, since the number of classifications of terminals is reduced, it is possible to simplify the determination algorithm of the connection destination (transfer source) terminal.

(C) Other Embodiments In the description of each of the above-described embodiments, various modified embodiments have been referred to. However, modified embodiments as exemplified below can be given.

  In each of the above embodiments, the node that transfers files is a terminal, but the present invention is not limited to a terminal. That is, it is only necessary that the nodes that transfer files correspond to each other, and the file is not limited to a terminal. For example, in a system in which a broadcast key station device distributes content to local station devices under its control, and each station device distributes content to each user's device by wire, content to each station device The present invention can also be applied to the distribution of the above.

Further, in each of the above embodiments, a case where a file is transferred from one transfer source device to a file transfer destination terminal has been described. However, a file may be transferred from a plurality of transfer source devices to one transfer destination device. May be. For example, when a terminal B that is receiving a file from the terminal A is determined as a transfer source terminal to a certain terminal C, the terminal B receives the file portion that has been received when a transfer request from the terminal C is given. May be transferred to the terminal C, and the terminal A may copy the remaining file portion to be transferred to the terminal B in accordance with an instruction from the terminal B and transfer it to the terminals B and C in parallel.

  In the above-described embodiment, the priority is given to the terminal of the higher layer among the receiving terminals. However, the terminal with the earlier inquiry time or reception start time may be prioritized.

  In addition, the receiving terminal may not be recognized as the transfer source terminal. In this case, the terminal with the latest reception completion may be given priority in determining the connection destination terminal. Further, when the number of nodes constituting the system is small, each node may be able to transfer one file only to one other node. Note that, even if the terminal being received is included, priority may be given to determining the connection destination terminal that has a later reception start time.

  In the above description, the file to be transferred is described as an image of moving image content, but the type of transfer file is not limited.

It is a block diagram which shows the logical function structure of the index server and client terminal in 1st Embodiment. It is a block diagram which shows the outline | summary of the network connection of the component of the file transfer system which concerns on 1st Embodiment. It is explanatory drawing of the storage content of the terminal database part in 1st Embodiment. It is explanatory drawing of the structure of the upload list in 1st Embodiment. It is a flowchart which shows the operation | movement at the time of the system connection of the client terminal in 1st Embodiment. It is a flowchart which shows the operation | movement by the side of a client terminal when the client terminal in 1st Embodiment receives the transfer of the desired file. It is a flowchart which shows the operation | movement by the side of an index server in case the client terminal in 1st Embodiment receives the transfer of the desired file. It is explanatory drawing of the specific determination method of the connecting point terminal (transfer origin terminal) in 1st Embodiment. It is explanatory drawing of the characteristic of the file transfer system in 2nd Embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... File transfer system, 11 ... Internet, 12 ... Index server, 13 ... Client terminal, 21 ... Control part, 22 ... Terminal selection part, 23 ... Information correction part, 24 ... Log recording / storage part, 25 ... Input / output part , 26 ... terminal database part, 27 ... file database part, 31 ... control part, 32 ... setting storage part, 33 ... log recording / storage part, 34 ... file storage part, 35 ... input / output part, 36 ... partial cache storage part 37 ... Connection state monitoring unit.

Claims (5)

File transfer in which a node that wants a file queries the index server for the node that is the transfer source node holding the file, receives a reply from the transfer source node determined by the index server, and receives a file transfer from the transfer source node In the system,
The index server
Topology information management means for managing, for each file , topology information of a tree structure in which transfer sources and transfer destination nodes in the past transfer are connected by links having directions;
Parameters for determining a transfer source node other than in the topology structure, and for each node that can be a transfer source node, a parameter management unit for determination that manages at least a region to which the node belongs and an available bandwidth,
A transfer source node determining means for determining the transfer source node,
Upper Symbol forwarding node determining means from the candidates of the source node, a large node bandwidth usage is a node belonging to the neighborhood local node wants the file, and determines on the forwarding node File transfer system. The transfer source node determination means determines a transfer source node as a transfer source node candidate for a receiving node receiving a file transfer from a certain node, and the transfer source node for which transfer has been completed 2. The file transfer system according to claim 1, wherein a candidate of a transfer source node being received is determined as a transfer source node only when no candidate exists . File transfer in which a node that wants a file queries the index server for the node that is the transfer source node holding the file, receives a reply from the transfer source node determined by the index server, and receives a file transfer from the transfer source node In the index server in the system,
Topology information management means for managing, for each file , topology information of a tree structure in which transfer sources and transfer destination nodes in the past transfer are connected by links having directions;
Parameters for determining a transfer source node other than in the topology structure, and for each node that can be a transfer source node, a parameter management unit for determination that manages at least a region to which the node belongs and an available bandwidth,
A transfer source node determining means for determining the transfer source node,
Upper Symbol forwarding node determining means from the candidates of the source node, a large node bandwidth usage is a node belonging to the neighborhood local node wants the file, and determines on the forwarding node Index server. File transfer in which a node that wants a file queries the index server for the node that is the transfer source node holding the file, receives a reply from the transfer source node determined by the index server, and receives a file transfer from the transfer source node In the method
The index server includes topology information management means, determination parameter management means and transfer source node determination means,
The topology information management means connecting the transfer source and the transfer destination node in a past transfer link having a direction, the topology information of the tree structure, and managed for each file,
The determination parameter management means is a parameter for determining a transfer source node other than the topology structure, and for each node that can be a transfer source node, manages at least a region to which the node belongs and an available bandwidth,
The transfer source node determining means determines, from the transfer source node candidates, a node belonging to an area near the node that desires the file and having a large use band as the transfer source node. Transfer method. File transfer in which a node that wants a file queries the index server for the node that is the transfer source node holding the file, receives a reply from the transfer source node determined by the index server, and receives a file transfer from the transfer source node In the file transfer program installed in the index server of the system,
Topology information management means for managing, for each file , topology information of a tree structure in which transfer sources and transfer destination nodes in the past transfer are connected by links having directions;
Parameters for determining a transfer source node other than in the topology structure, and for each node that can be a transfer source node, a parameter management unit for determination that manages at least a region to which the node belongs and an available bandwidth,
A node that belongs to the neighborhood of the node that wants the file and that has a large bandwidth is made to function as a transfer source node determination unit that determines the transfer source node from among the transfer source node candidates. File transfer program.

Images