JP5228369B2 - COMMUNICATION SYSTEM, COMMUNICATION METHOD, AND COMMUNICATION PROGRAM - Google Patents

Description

  The present invention relates to a communication system, a communication method, and a program, and more particularly, to a communication system, a communication method, and a program in which terminals communicate directly with each other without using a server on an IP network.

  In recent years, not only text but also various contents such as images, sounds, and videos have been directly exchanged between users by a file sharing network using P2P (Peer to Peer) technology.

  An example of the P2P technology is a code (Chord). A code (Chord) is one of the P2P technologies using a distributed hash table. Participating nodes are given IDs (identifiers) generated using a hash function, and are connected in a ring shape on the overlay network in the order of the IDs. . The content is similarly given an ID by a hash function, and the combination data of the content ID and the content location information is stored in a node participating in the network according to the ID. When searching for content with a node, a search message is routed between nodes using a distributed hash table to search for a node having location information of the desired content. Nodes maintain a chord ring by periodically checking their existence with neighboring nodes.

  A code (Chord) can also be used for the purpose of connecting terminals such as when making a phone call. For example, in P2P-SIP (Session Initiation Protocol), SIP-URI (Uniform Resource Identifier) is used as content, and the result of applying SIP-URI to the hash function is registered as a content ID in a code (Chord) ring together with IP (Internet Protocol) address. In the case of communication, the IP address is searched on the code ring using the SIP-URI as a key.

  In a general P2P network as well as a code (Cord), a participating node can leave the P2P network at an arbitrary timing. In the case of a code (Cord), if a node leaves, the content ID and the IP address information disappear, so that information needs to be backed up.

  For example, there is a method in which a copy of information is transferred to several nodes upstream and downstream along the ring.

  Alternatively, as disclosed in Patent Document 1 (Japanese Patent Application Laid-Open No. 2004-72603), there is a method in which a backup node is set for each node and a copy of data is transferred to the node.

Thus, by leaving the combination data of the content ID and the content location information, the code (Chord) ring can be operated normally even if the node is detached.
JP 2004-72603 A

  The conventional problem is that even if a node leaves, the information of the node must be backed up to another node in order to operate normally.

  This is because a node is allowed to leave at an arbitrary timing in a P2P network. In order to do so, a node must copy its own information before leaving another node. Also, both nodes and backup nodes may leave the P2P network at the same time, so to avoid such a situation, the number of backup nodes must be increased and as a result must be stored in one node. The amount of information that must be handled becomes large. That is, in order to increase the ring repair capability, the amount of information that one node must store increases, while on the other hand, if the amount of information that one node must store is reduced, the ring repair capability decreases.

(the purpose)
An object of the present invention is to provide a communication system capable of repairing a P2P network when any node participating in the P2P network leaves without giving a node redundant information.

  In order to solve the above problems, the present invention provides a content-oriented P2P network in which unique identification information is assigned to the content on a node holding one or a plurality of content and the content is the content node. It is characterized by constituting a ring.

  In the following, means for solving the problem will be described in more detail using the numbers used in the best mode for carrying out the invention in parentheses. These numbers are added to clarify the correspondence between the description of the claims and the best mode for carrying out the invention. However, these numbers should not be used for the interpretation of the technical scope of the invention described in the claims.

  The first communication system of the present invention includes a plurality of nodes having one or a plurality of contents, and each node is connected to an external network, and a transmission / reception unit 1 that transmits / receives information to / from other nodes or content nodes; One content is created for each content registered in the content database 6 and added when the content is registered in the content database 6 and the content node function unit 2 that provides the content node function for the content in charge. Content node generation function unit 3 for adding content node function unit 2 in charge of content, and content node deletion for deleting content node function unit 2 in charge of deleted content when content is deleted from content database 6 Function part 4 and the controller you want to search The content node search function unit 5 for searching for content and the content database 6 for storing the content to be disclosed to other nodes by obtaining the ID number of the content from the hash function and transmitting the content search request to the content node function unit And a finger table database 7 for storing a finger table used in the content node function unit 2 existing in the node. Adopting such a configuration, making the P2P network a content-oriented configuration, and content nodes participating in the network eliminates the need to register content location information in each node, and a certain node leaves. When a content node leaves the network, the content node on the downstream side of the detached content node searches for the upstream content node of the detached content node, and combines the two to repair the network. Can be achieved.

  The transmission / reception unit 1 is connected to an external network, and transmits information on a DHT participation message, a content search message, and a keep alive message transmitted from the content node function unit 2 to other nodes described in the message. And a function for notifying the content node function unit 2 described in the message of the DHT participation message, the content search message, and the keep alive message from other nodes. Have. This operation realizes transmission / reception of information with other nodes or other content nodes.

  The content node function unit 2 is generated for each content registered in the content database 6, and a DHT participation function unit 21 that provides a function of registering the content node of the content in charge on the P2P network; A message transmission destination is determined according to the algorithm of DHT, a message transmission destination is described in the content search message, and a DHT search function unit 22 that transmits to the transmission / reception unit 1 transmits / receives a keep alive message to / from other content nodes. And a DHT management function unit 23 having a function of maintaining the network. With this configuration, a DHT management for configuring the network and a content search function on the network are provided.

  When the content node function unit 2 is generated by the content node generation function unit, the content node function unit 2 receives a DHT participation message from a function for causing the content node function unit 2 to participate in the P2P network using the DHT participation function unit 21 and the other content node. When a content search request is received from the content node search function unit 5 by using the DHT management unit 23 to receive the content node that has transmitted the DHT participation message on the network, the DHT search function unit 22 is used. The content table is searched according to a predetermined algorithm after referring to the finger table database 7 and referring to the finger table related to other content nodes existing in the same node other than the finger table related to the content node. A message transmission destination is determined, a message transmission destination is described in the content search message, and the message transmission destination is transmitted to the transmission / reception unit 1. When a content search message is received from another content node, the DHT search function unit is used to refer to the finger table database 7 The content node is searched according to a predetermined algorithm with reference to the finger table related to other content nodes existing in the same node other than the finger table related to the self content node, the content message transmission destination is determined, and the message is transmitted to the content search message. It has a function of transmitting the above to the transmission / reception unit 1 and a function of deleting itself upon receiving a content node deletion request from the content node deletion function unit 4.

  The DHT participation function unit 21 has a function of registering the content node of the content in charge on the P2P network by transmitting a DHT participation message to other content nodes.

  When the DHT search function unit 22 receives the content node search request from the content node search function unit, the DHT search function unit 22 refers to the finger table database 7 and refers to fingers related to other content nodes existing in the same node other than the finger table related to the self content node. The message transmission destination is determined according to a predetermined algorithm with reference to the table, the function of sending the message transmission destination to the transmission / reception unit 1 described in the content search message, and the reception of the content search message from another content node Refer to the table database 7 and refer to the finger tables related to other content nodes existing in the same node other than the finger table related to the self content node, and follow a predetermined algorithm. Decide message destination, having a function of transmitting to the transceiver unit 1 on the described message destination to the contents search message.

  The DHT management unit 23 transmits / receives a keep alive message including the finger table information related to the content node included in the finger table database 7 with other content nodes, updates the information in the finger table, and updates the P2P network. The function to maintain and keep alive message transmission / reception periodically. After sending the keep alive message, there is no reply for a certain period of time, and the keep alive message is retransmitted. If there is no reply, the content node is assumed to have left the network, and the function to detect the removal of another content node and the detection of the departure of a content node are detected. If the content node is downstream when viewed from its own content node, it waits for a certain period of time for a content search message with a ring repair option to search for the content node that has left the downstream node as viewed from the content node When it is detected that a certain content node has left, and the separated content node is upstream as viewed from its own content node, a ring repair option for searching for a detached content node using the DHT search function unit 22 is added. By waiting for a content search message with a function to repair the network by communicating with the content node existing at the destination of the detached content node and a ring repair option to search for the detached content node. A function to complete ring repair by sending and receiving keep alive messages to and from a content node that has sent a content search message with a ring repair option to search for a detached content node. If a content search message with a ring repair option for searching for a detached content node does not arrive after a certain time elapses, the finger table database 7 is referred to. If the connection is successful, the content node and the ring are repaired. If the connection is successful, the content node and the ring are repaired in order. When waiting for a content search message with a function to repeat the above operation until a connectable content node is found with reference to the finger table database 7 and a ring search option with a ring repair option to search for a detached content node When a content search message with a ring repair option for searching for a plurality of detached content nodes is received, if there is an efficient content search message transmission destination by referring to the finger table database 7 as in the case of a normal content search message , A function for transmitting a content search message to the content node, and a case where there is no efficient content search message transmission destination by referring to the finger table database 7, that is, a plurality of ring repair options If the content search message attached is the optimal recipient, the ID number of the content search source of the content search message is compared, and the closest ID number that is larger than the ID number of the content search message is determined. It tries to connect in order from the content node that has, and after sending a content search message with a function to discard all other content search messages and a ring repair option to search for a detached content node, search for the detached content node A function that waits for a keep alive message from a content node waiting for a content search message with a ring repair option and a container with a ring repair option that searches for a detached content node. If a keep alive message is not received even after a certain time has elapsed since the transmission of the message search message, a function for retransmitting the content search message with a ring repair option for searching for a detached content node, When a keep alive message is not received for all the content search messages included, the content node having the closest ID number that is smaller than the ID number of its own content node is referred to the finger table database 7 If the connection is successful, the content node and the ring are repaired. If the connection is not successful, the finger table database 7 is referred again, and the above operation is repeated until a connectable content node is found. And having a function.

  The first communication method of the present invention periodically transmits and receives keep alive messages to and from other content nodes, and after sending a keep alive message, there is no reply for a certain period of time. If there is no reply even after resending the keep alive message, the content node detects that the content node has left the network by leaving the P2P network, and the network of the content node that has left the network Step S-2 for confirming the direction above, and Step S- for defining the operation when the direction of the content node that has left the network is downstream as viewed from the content node. A and the corresponding node of the detached content node Step SB that defines the operation when the content node that has left the network is upstream when viewed from its own content node, and a keep alive message are sent and received to repair the network. Step S-C1.

  Further, the step S-A waits for a certain period of time for a content search message with a ring repair option for searching for a detached content node from a downstream node as seen from the detached content node. Step S-A2 for confirming whether or not a content search message with a ring repair option for searching for a content node that has left the downstream node when viewed from the content node is received, and from the content node that has left When a content search message with a ring repair option for searching for a content node that has left the node that has been viewed is not received within a certain period of time, the finger table database is referred to and the ID number of the content node is Big and closest A step S-A3 that tries to connect in order from the content node having the D number, and a content search message with a ring repair option for searching for a detached content node from the downstream node as seen from the detached content node within a certain time If received, step S-A4 confirms whether there is an efficient content search message transmission destination by referring to the finger table database, and a keep alive message is sent to the content node determined in step S-A4. Transmitting step S-A6.

  Step S-B is a step S-B1 for transmitting a content search message with a ring repair option for searching for a detached content node, and a step S for confirming whether a keep alive message has been received within a predetermined time. Step S-B3 for resending a content search message with a ring repair option for searching for a detached content node, and step S-B4 for confirming whether or not a keep alive message has been received within a predetermined time. And step S-B5 which refers to the finger table database and tries connection in order from the content node having the closest ID number which is smaller than the ID number of the content node.

  The communication program of the present invention is a program for causing a computer to execute the communication method.

  The first effect of the present invention is that a network can be repaired when an arbitrary node that has participated in the P2P network leaves without giving a node redundant information.

  The reason for this is that a unique identification information is given to the content on a node that holds one or a plurality of contents, and a content-oriented P2P type network ring is configured with the content as a content node. is there. That is, a P2P type network ring is configured in a content-oriented manner, and content nodes participate in the network ring, so that it is not necessary to register content location information in each node. When the network node leaves the network ring, the content node on the downstream side of the content node that has left the network searches the upstream content node of the content node that has left the network and connects the two to restore the network.

(First embodiment)
A first embodiment of the present invention will be described below with reference to the accompanying drawings.

(Configuration of the first embodiment)
FIG. 1 is a schematic diagram showing a physical configuration of a communication system according to the present embodiment.

  Referring to FIG. 1, in the communication system according to the present embodiment, a node 100 that is an information processing terminal arranged on a network can have a plurality of contents.

  FIG. 2 is a schematic diagram showing a logical configuration of the communication system according to the present embodiment.

  Referring to FIG. 2, the communication system according to the present embodiment configures a P2P (Peer to Peer) network in a content-oriented manner, and assigns a unique node ID (identifier) to each content (content node) of the node 100. Allocation, node IDs (identifiers) 2, 3,..., 56, 61 are made to participate in the structured network. Note that the topology of the network according to the present embodiment is a ring (ring network), but the topology is not limited to a ring.

  FIG. 3 is a block diagram showing a configuration of a node according to the present embodiment.

  Referring to FIG. 3, the nodes constituting the communication system according to the present embodiment are the transmission / reception unit 1, content node function unit 2, content node generation function unit 3, content node deletion function unit 4, and content node search. It has a function unit 5, a content database 6, and a finger table database 7.

  The content node function unit 2 includes a DHT (Distributed Hash Table) participation function unit 21, a DHT search function unit 22, and a DHT management function unit 23.

  The transmission / reception unit 1 is connected to an external network, and transmits information on a DHT participation message, a content search message, and a keep alive message transmitted from the content node function unit 2 to other nodes described in the message. Send. When a DHT join message, a content search message, and a keep alive message are received from another node, the message is notified to the content node function unit 2 described in the message.

  One content node function unit 2 is generated for each content registered in the content database 6, and provides a content node function for the content in charge. That is, one of the content node function units 2 corresponds to one content node on the ring network, and provides a DHT management for configuring the ring network and a content search function on the ring network. Here, the ID (identifier) on the ring network of the content node uses a result of applying information on the content to the hash function. The information regarding the content includes, for example, a content name.

  When the content node function unit 2 is generated by the content node generation function unit 3, the content node function unit 2 uses the DHT participation function unit 21 to join the own content node (own content node function unit 2) on the ring network.

  When the content node function unit 2 receives the DHT participation message from another content node, the content node function unit 2 uses the DHT management unit 23 to cause the content node that has transmitted the DHT participation message to participate in the ring network.

  When the content node function unit 2 receives a content search request from the content node search function unit 5, the content node function unit 2 transmits a content search message to the transmission / reception unit 1 using the DHT search function unit 22. That is, the DHT search function unit 22 refers to the finger table database 7, determines a message transmission destination according to a predetermined algorithm described later, and transmits the message search destination to the transmission / reception unit 1 with the message transmission destination described in the content search message.

  Here, the finger table database 7 stores finger tables related to all the content node function units 2 existing in the node. When the DHT search function unit 22 refers to the finger table database 7 and determines a message transmission destination according to a predetermined algorithm, the DHT search function unit 22 refers to a finger table related to another content node function unit in addition to the finger table related to the own content node function unit. It is possible to obtain content node information that cannot be known only by referring to the finger table related to the content node. That is, since a lot of content node information can be referred to, an efficient message transmission destination can be determined. Here, the efficient message transmission destination is the content node itself to be searched by the content search message. If there is no information, the content node has an ID number smaller than the ID number of the content node to be searched. The content node with the largest ID number.

  When the content node function unit 2 receives a content search message from another content node, the content node function unit 2 searches the content node according to a predetermined algorithm using the DHT search function unit 22 and the finger table database 7, determines the content message transmission destination, The message transmission destination is described in the search message, and is transmitted to the transmission / reception unit 1.

  Further, upon receiving the content node deletion request from the content node deletion function unit 4, the content node function unit 2 deletes itself.

  The DHT participation function unit 21 provides a function of registering the content node of the content in charge on the ring network. That is, in order to register a content node on the ring network, a DHT participation message is transmitted to other content nodes.

  Upon receiving the content node search request from the content node search function unit 5, the DHT search function unit 22 refers to the finger table database 7, determines a message transmission destination according to a predetermined algorithm, and describes the message transmission destination in the content search message. To the upper transmission / reception unit 1.

  When the DHT search function unit 22 receives a content search message from another content node, the DHT search function unit 22 refers to the finger table database 7, determines a message transmission destination according to a predetermined algorithm, and describes the message transmission destination in the content search message. Calls to the transceiver 1.

  The DHT management unit 23 has a function of transmitting / receiving keep alive messages to / from other content nodes and maintaining a ring network. Here, the keep alive message also includes a finger table related to its own content node included in the finger table database 7, and the finger table is updated by exchanging the finger table with other content nodes. . That is, when the finger table received from another content node is newer than the finger table of the content node, the information of the finger table received from the other content node is referred to update the finger table of the content node. Here, a keep alive message is executed for one or more content nodes upstream and downstream on the ring network.

  Here, upstream on the ring network refers to the direction in which a content node with a small ID number exists when viewed from a content node on the ring network, and downstream on the ring network refers to a content node on the ring network. This refers to the direction in which content nodes with large ID numbers exist. However, if the content node ID in either direction as viewed from a certain content node is small in the portion corresponding to the joint of the ID number, the content node side having the smaller ID number is set as the downstream side, and the opposite side is set as the upstream side.

  Furthermore, the DHT management unit 23 detects the detachment of other content nodes by periodically transmitting and receiving keep alive messages. That is, after sending a keep alive message, if there is no reply for a certain period of time and there is no reply even if the keep alive message is retransmitted, it is assumed that the content node has left the ring network.

  When the DHT management unit 23 detects that a certain content node has left, the DHT management unit 23 has a function of communicating with the content node existing at the detached content node and repairing the ring network. When the detached content node is downstream as viewed from the content node, the content search message with a ring repair option for searching for the detached content node from the downstream node as viewed from the detached content node is waited for a certain period of time. On the other hand, if the detached content node is upstream from the content node, a content search message with a ring repair option for searching for the detached content node using the DHT search function unit 22 is transmitted.

  The DHT management unit 23 waiting for the content search message with the ring repair option for searching for the detached content node keeps the content node that has transmitted the content search message with the ring repair option for searching for the detached content node. Ring repair is completed by sending and receiving alive messages.

  Here, when waiting for a content search message with a ring repair option to search for a detached content node, content search with a ring repair option to search for a detached content node depending on the state of the network and the damage state of the ring network It is also possible that the message does not arrive after a certain time. In this case, referring to the finger table database 7, connection is attempted in order from the content node having the closest ID number that is larger than the ID number of the self-content node. If the connection is successful, the content node and the ring are repaired. If the connection is not successful, the finger table database 7 is referred again, and the above operation is repeated until a connectable content node is found.

  In addition, when waiting for a content search message with a ring repair option to search for a detached content node, it is possible to receive a content search message with a ring repair option to search for multiple detached content nodes within a certain period of time. There is also sex. In this case, first, when there is an efficient content search message transmission destination by referring to the finger table database 7 as in the case of a normal content search message, the content search message is transmitted to the content node. On the other hand, when there is no efficient content search message transmission destination with reference to the finger table database 7, that is, when the content node is the optimum recipient for the content search message with a plurality of ring repair options, the content search message is transmitted. The ID numbers of the original content nodes are compared, and connection is attempted in order from the content node having the closest ID number that is larger than the ID number of the self content node. All other content search messages are discarded.

  When a content search message with a ring repair option for searching for a detached content node is transmitted, the DHT search function unit 22 refers to the finger table database 7 in the same manner as a content search message without a ring repair option. By referring to the finger table related to the other content node function unit 2 in addition to the finger table related to the own content node function unit 2, it is possible to obtain content node information that cannot be known only by referring to the finger table related to the own content node. . That is, since a lot of content node information can be referred to, an efficient message transmission destination can be determined. Here, an efficient message transmission destination is a content node that has the largest ID number among content nodes that have an ID number smaller than the ID number of the content node that is to be searched by the content search message, that is, the detached content node. . By repeating such a search, a message is delivered to a content node waiting for a content search message with a ring repair option for searching for a detached content node.

  After sending a content search message with a ring repair option to search for a detached content node, the DHT manager 23 keeps the content node waiting for the content search message with a ring repair option to search for a detached content node. Wait for a keep alive message.

  Here, if a keep alive message is not received even after a certain time has elapsed, a content search message with a ring repair option for searching for a detached content node is retransmitted. The number of retransmissions is not limited to one, and retransmissions can be performed a plurality of times depending on the network state.

  If a keep alive message is not received for all of the resent content search messages, the finger table database 7 is referred to, and the ID number smaller than the ID number of the content node is the closest. Try to connect in order from the content node. If the connection is successful, the content node and the ring are repaired. If the connection is not successful, the finger table database 7 is referred again, and the above operation is repeated until a connectable content node is found.

  When the content is registered in the content database 6, the content node generation function unit 3 adds the content node function unit 2 in charge of the added content.

  When the content is deleted from the content database 6, the content node deletion function unit 4 deletes the content node function unit 2 that was responsible for the deleted content.

  That is, when a new content is added to the content database 6, the node uses the content node generation function unit 3 to add the content node function unit 2 in charge of the added content. The added content node function unit uses the DHT participation function unit 21 to participate in the ring network. On the other hand, when content is deleted from the content database 6, the node uses the content node deletion function unit 4 to delete the content node function unit 2 that was responsible for the deleted content.

  The content node search function unit 5 searches the content by obtaining the ID number of the content to be searched from the hash function and transmitting a content search request to the content node function unit 2. Here, among the content node function units 2 existing in the node, the content node function unit 2 in charge of the largest ID number among the content node function units in charge of content whose ID number is smaller than the ID number of the content to be searched. On the other hand, an efficient search can be implemented by transmitting a content search request.

  The content database 6 stores content that is disclosed to other nodes. Note that content is a concept provided for convenience, and includes information indicating the location of an information resource on the Internet, such as a URI (Uniform Resource Identifier).

  The finger table database 7 stores a finger table used in the content node function unit 2 existing in the node. In addition, when a plurality of finger tables exist in the node 100, the node 100 associates each finger table and stores them in the finger table database 7. The finger table is a routing table used when searching for a content node in the predetermined algorithm in the present embodiment, and describes information related to the content node such as the ID number and IP address of the content node.

A hash space is defined with a value of 2 ^m (m is a parameter), and a node ID for identifying each content node is calculated and assigned to each content node of the node 100 using a hash function.

  Accordingly, one node 100 may have a plurality of node IDs, and when one node 100 has a plurality of contents (node IDs), it appears that a plurality of nodes exist by each content node in the hash space. .

  A set of an IP address that is the location of the node 100 and a node ID that is the location of the content is recorded in the finger table database 7 as a finger table, and the information of the finger table is stored in each of the predetermined algorithms in the present embodiment. Distribute to nodes.

  Each content node has a node ID and an IP address of a content node (P node) located next to the upstream side of the content node on the hash space, and content located next to the downstream side of the content node on the hash space It has a node ID and an IP address of the node (S node).

Also, each content node has m entries of a set of node ID and IP address as a finger table, and its own node ID = A, and the node ID of the i-th entry is (A + 2 ⁱ ) mod 2 ^m 0 ≦ i When it is determined by the expression (1) of ≦ m−1, the node ID of the existing node having a node ID larger than the node ID of the i-th entry is included.

  FIG. 4 is a diagram collectively showing the finger table database 7 included in each node according to the embodiment.

Referring to FIG. 4, in the hash space, each content node has a node ID of a content node (P node) located next to the upstream side of the content node, and a content node located next to the downstream side of the content node. The node ID of (S node) and the node IDs from m = 0 to m = 6-1 determined by the above equation (1) in the hash space 2 ⁶ (m = 6).

  For example, the content node of the node ID has P = 61, S = 3, i (0) = 3, i (1) = 8, i (2) = 8, i (3) = 13, i as finger tables. (4) = 18 and i (5) = 37 information.

  The present embodiment can also be applied to a structured ring network to which a unique ID is assigned without using a distributed hash.

(Operation of the first embodiment)
Next, the operation of the communication system when the hash space of the present embodiment is 2 ⁶ = 64 will be described in detail with reference to the drawings.

  FIG. 5 is a sequence diagram showing a stabilization operation according to the present embodiment.

  Each content node periodically sends and receives keepalive messages to and from the P node (predecessor node) and to the S node (successor node), and exchanges difference data with each other.

  Each content node provides a data difference by using a flag indicating which data is new to the party performing the stabilization operation, and if there is no data difference (data has not been updated) Case), perform an empty response.

  If there is no response to the stabilization message requesting the data difference, each content node uses the stored data and uses the stored data to send a stabilization message to the S node adjacent to the S node that does not respond. Is transmitted, or a stabilization message is transmitted to the P node adjacent to the P node that does not respond, and the stabilization operation is performed.

  FIG. 6 is a sequence diagram showing a stabilization operation when a node or a content node leaves the ring network.

  Referring to FIG. 6, the content node with node ID 40 and the content node with node ID 48 periodically send and receive a stabilization message, and the content node with node ID 33 and the content node with node ID 40 stay with each other. It regularly sends and receives the visualization message.

  Next, the content node of the node ID 40 notifies the content node of the node ID 48 and the content node of the ID 33 that regularly send and receive the stabilization message to each other without notifying that the ring network is to be disconnected. When leaving the node, the content node node with the node ID 48 and the content node with the ID 33 recognize the detachment of the content node with the node ID 40 because there is no response to the transmitted stabilization message.

  Next, the content node with the node ID 33 transmits a stabilization message to the S node (content node with the node ID 48) adjacent to the S node (content node with the node ID 40) that does not respond, while the node ID 48 has the node ID 48. The content node transmits a stabilization message to the P node (content node of node ID 33) adjacent to the P node (content node of node ID 40) that does not respond, and periodically sends and receives the stabilization message to each other. Do it.

  Note that the content node with the node ID 40 notifies the content node with the node ID 48 and the content node with the ID 33 that regularly send and receive a stabilization message to the effect that they leave the ring network (end message). At the time of the normal termination after leaving, the operation of the node (the content node with node ID 33 and the content node with node ID 48) that has received the end message is the same as that at the time of leaving recognition in the stabilization operation.

  Next, an outline of basic operations in ring network restoration will be described.

  First, when the node 100 or the content node detects the disappearance of two nodes (content nodes) with consecutive node IDs (withdrawal from the ring network), the node 100 or the content node from which the upstream node has left determines the transmission destination of itself. As a node having the largest node ID value on the finger table, a message for searching for a node (content node) with an error message is issued. The node 100 or content node from which the downstream node has left shifts to an error waiting state.

  Each node on the ring network transfers messages in a normal sequence.

  When receiving the message, the node 100 or the content node in the error waiting state compares the search destination with the finger table, and the node (content node) to which the maximum node ID that does not exceed the value of the node ID of the search destination is given. Forward the received message.

  When receiving a message, the node 100 or the content node waiting for an error is connected to the message transmission source when there is no forwarding destination (= the node (content node) that satisfies the condition).

  Note that when receiving a plurality of messages, the node 100 or the content node waiting for an error gives priority to transferring the message with the smaller node ID.

  On the other hand, the node 100 or the content node from which the upstream node has left resends the message only once when the response to the transmitted message does not wait for a certain time.

  FIG. 7 is a flowchart showing a more detailed operation example of the ring network repair when the node 100 or the content node leaves the ring network.

  Referring to FIG. 7, when the node 100 or the content node detects the disappearance (leaving from the ring network) of two nodes (content nodes) having consecutive node IDs (step S-1), the disappeared node (content node) The node ID is compared with the value of its own node ID to determine the magnitude of the node ID (step S-2).

  In step S101, the node 100 or the content node periodically detects the detachment of the other content node by periodically transmitting and receiving a keep alive message with the other content node. That is, after sending a keep alive message, if there is no reply for a certain period of time and there is no reply even if the keep alive message is retransmitted, it is assumed that the content node has left the ring network.

  When the removal of another content node is detected, the operation differs depending on the direction of the detached content node on the ring network. When the detached content node is downstream as viewed from the own node 100 or the content node, the process proceeds to step S-A1. On the other hand, when the detached content node is upstream from the own node 100 or the content node, the process proceeds to step S-B1.

  When the node ID of the extinguished (leaved) node (content node) is smaller than its own node ID (when the upstream node leaves), the node 100 or the content node that detected the leaving of the node (content node) A search message with an error with a ring repair option for searching for a node is awaited (step S-A1).

  Next, when the node 100 or the content node that has detected the withdrawal of the node (content node) does not receive the search message with an error within a certain time (step S-A2), the finger table database is referred to The above-mentioned stabilization operation is performed with a node (content node) having a node ID closest to its own node ID (step S-A3).

  That is, in step S-A3, the node 100 or the content node that has detected the departure of the node (content node) tries to connect in order from the content node having the closest node ID that is larger than the node ID of the content node, If the connection is successful, the process proceeds to step S114. If the connection is not successful, the finger table database is referred to again. The above operation is repeated until a connectable content node is found, and then the process proceeds to step S114.

  On the other hand, when the node 100 or the content node that detects the withdrawal of the node (content node) receives the search message with an error within a certain time (step S-A2), it determines whether or not it is the optimum recipient. (Step S-A4). In this determination, when the maximum node ID value that does not exceed the maximum value of the node ID to be searched is the own node ID, it is determined that the node is the optimum recipient.

  In step S-A4, when the node 100 or the content node that has detected the withdrawal of the node (content node) determines that it is not the optimum recipient, it forwards the search message with an error based on the finger table (step S-A). A5) Returning to step S-A2, on the other hand, if it is determined that it is the optimum recipient, a response (a keep alive message) is transmitted in response to the search message with an error (step S-A6). .

  In step S-A6, when the node 100 or the content node that detected the withdrawal of the node (content node) receives a plurality of search messages with errors, the error received from the transmission source having a smaller node ID of the transmission source Priority is given to sending a response to a search message with an error, and a search message with an error received from another is discarded.

  Next, the node 100 or the content node that has detected the detachment of the node (content node) is connected to the response transmission source node (content node) to transmit and receive a keep alive message (step S-C1). ), The ring network is repaired.

  On the other hand, in step S-2, when the node ID of the extinguished (leaved) node (content node) is larger than its own node ID (when the downstream node leaves), the node (content node) is detected to leave. The node 100 or the content node transmits a search message with an error with a ring repair option for searching for the detached node (content node) by setting the search target as the node (content node) (step S-B1).

  Next, the node 100 or the content node that has detected the withdrawal of the node (content node) receives a response (keep alive message) to the transmitted search message with an error within a certain time (step S-B2). The ring network is repaired by connecting to the node (content node) that sent the response and transmitting / receiving a keep alive message (step S-C1).

  On the other hand, in step S-B2, the node 100 or the content node that has detected the withdrawal of the node (content node) does not receive a response to the transmitted search message with an error (keep alive message). After a lapse of time, a search message with an error with a ring repair option for searching for a detached node (content node) is retransmitted (step S-B3), and a response to the resent search message with an error (keep alive) message ) Is not received within a certain time (step S-B4), the above-described stabilization operation is performed with the node (content node) having the node ID farthest from its own node ID on the finger table (step S-). B5). It should be noted that the node 100 or the content node that has detected the withdrawal of the node (content node) goes down when the stabilization operation cannot be performed in step S-B5.

  On the other hand, if a response (keep alive message) to the resent search message with an error is received within a certain time (step S-B4), the connection is made with the response transmission source node (content node) and keep alive. By transmitting and receiving (keep alive) messages (step S-C1), the ring network is restored.

  In step S-2, when the upstream node and the downstream node leave, both the processing of step S-A1 and the processing of step S-B1 are performed.

(Effects of the first embodiment)
Next, the effect of this embodiment will be described.

  According to the present embodiment, a P2P network is configured in a content-oriented manner, a unique node ID is allocated to content (content node) on a node that can have a plurality of contents, and the structured network is joined to the network. This eliminates the need to register content location information in each node, and when a certain node leaves and the content node leaves the network, the content node downstream of the detached content node By searching the upstream content node, connecting the two and repairing the network, the ring can be repaired when an arbitrary node leaves without redundant information.

  Further, according to the present embodiment, since the finger table relating to all the content node function units 2 existing in one node 100 is stored in the finger table database 7 of the node 100, the DHT search function unit 22 When the message transmission destination is determined according to a predetermined algorithm with reference to the database 7, the finger table related to the content node function unit 2 is also referred to in addition to the finger table related to the content node function unit 2, thereby It is possible to obtain content node information that cannot be found only by referring to the table. That is, since a lot of content node information can be referred to, an efficient message transmission destination can be determined. Therefore, it is possible to prevent the message from being exchanged between a plurality of content node function units 2 existing in one node 100, and the message once output from a certain node 100 returns to the node 100. Can be prevented.

(Second Embodiment)
Hereinafter, a second embodiment of the present invention will be described with reference to the accompanying drawings.

  The network of the communication system according to the present embodiment is different from the first embodiment in that the network of the communication system is a chord ring using chord technology. The following description will focus on the points that differ from the first embodiment.

  (Configuration of Second Embodiment)

  One content node function unit 2 according to the present embodiment is generated for one content registered in the content database 6 and provides a content node function for the content in charge. That is, one of the content node function units 2 corresponds to one content node on the code ring, DHT management for configuring the code ring, and content search function on the code ring I will provide a. Here, the ID on the code ring of the content node uses the result of applying information about the content to the hash function. The information regarding the content includes, for example, a content name.

  When the content node function unit 2 according to the present embodiment is generated by the content node generation function unit 3, the content node function unit 2 uses the DHT participation function unit 21 to place its own content node (content node function unit 2) on the code ring. Let them participate.

  When the content node function unit 2 according to the present embodiment receives a DHT participation message from another content node, the content node function unit 2 uses the DHT management unit 23 to select the content node that has transmitted the DHT participation message on the code ring. To join.

  When the content node function unit 2 according to the present embodiment receives a content search request from the content node search function unit 5, the content node function unit 2 transmits a content search message to the transmission / reception unit 1 using the DHT search function unit 22. That is, the DHT search function unit 22 refers to the finger table database 7, determines a message transmission destination in accordance with a code algorithm, and transmits the message transmission destination to the transmission / reception unit 1 with the message transmission destination described in the content search message.

  Here, the finger table database 7 according to the present embodiment stores finger tables relating to all the content node function units 2 existing in the node. When the DHT search function unit 22 refers to the finger table database 7 and determines a message transmission destination according to a code (Chord) algorithm, the finger table related to other content node function units is also referred to in addition to the finger table related to the own content node function unit. Thus, it is possible to obtain content node information that cannot be known only by referring to the finger table related to the content node. That is, since a lot of content node information can be referred to, an efficient message transmission destination can be determined. Here, the efficient message transmission destination is the content node itself to be searched by the content search message. If there is no information, the content node has an ID number smaller than the ID number of the content node to be searched. The content node with the largest ID number.

  When the content node function unit 2 according to the present embodiment receives a content search message from another content node, the content node function unit 2 searches the content node according to a code algorithm using the DHT search function unit 22 and the finger table database 7, A message transmission destination is determined, and the message transmission destination is described in the content search message and transmitted to the transmission / reception unit 1.

  The DHT participation function unit 21 according to the present embodiment provides a function of registering the content node of the content in charge on the code (Chord) ring. That is, in order to register a content node on the code (Chord) ring, a DHT participation message is transmitted to other content nodes.

  Upon receiving the content node search request from the content node search function unit 5, the DHT search function unit 22 according to the present embodiment refers to the finger table database 7, determines a message transmission destination according to a code (Chord) algorithm, and searches for content The message transmission destination is described in the message, and is transmitted to the transmission / reception unit 1.

  Further, when the DHT search function unit 22 according to the present embodiment receives a content search message from another content node, the DHT search function unit 22 refers to the finger table database 7 and determines a message transmission destination according to a code (Chord) algorithm. The message transmission destination is described and transmitted to the transmission / reception unit 1.

  The DHT management unit 23 according to the present embodiment has a function of transmitting and receiving keep alive messages with other content nodes and maintaining a chord ring. Here, the keep alive message also includes a finger table related to its own content node included in the finger table database 7, and the finger table is updated by exchanging the finger table with other content nodes. . That is, when the finger table received from another content node is newer than the finger table of the content node, the information of the finger table received from the other content node is referred to update the finger table of the content node. Here, the keep alive message is executed for one or more content nodes upstream and downstream on the chord ring.

  Here, upstream on the code (Chord) ring indicates a direction in which a content node having a small ID number as seen from a content node in the code (Chord) ring exists, and downstream on the code (Chord) ring is This indicates a direction in which a content node having a large ID number as seen from a certain content node exists in a code ring. However, if the content node ID in either direction as viewed from a certain content node is small in the portion corresponding to the joint of the ID number, the content node side having the smaller ID number is set as the downstream side, and the opposite side is set as the upstream side.

  Furthermore, the DHT management unit 23 according to the present embodiment detects the detachment of other content nodes by periodically transmitting and receiving a keep alive message. In other words, after sending a keep alive message, if there is no reply for a certain period of time and there is no reply even if the keep alive message is retransmitted, it is assumed that the content node has left the chord ring. .

  When the DHT management unit 23 according to the present embodiment detects that a certain content node has left, the DHT management unit 23 has a function of communicating with the content node existing at the detached content node destination and repairing the code (Chord) ring.

  When new content is added to the content database 6, the node 100 according to the present embodiment uses the content node generation function unit 3 and adds the content node function unit 2 in charge of the added content. The added content node function unit according to the present embodiment uses the DHT participation function unit 21 to participate in the code ring. On the other hand, when content is deleted from the content database 6, the node uses the content node deletion function unit 4 to delete the content node function unit 2 that was responsible for the deleted content.

  The finger table database 7 according to the present embodiment stores a finger table used in the content node function unit 2 existing in the node. The finger table is a routing table used when searching for a content node in the code (Chord) algorithm, and describes information related to the content node such as an ID number and an IP address of the content node.

(Operation of Second Embodiment)
Next, the operation of the present embodiment will be described in detail with reference to FIG.

(1) Step S-11
It is an initial state of the present invention.

(2) Step S-12
By periodically sending and receiving keep alive messages with other content nodes, the detachment of other content nodes is detected. In other words, after sending a keep alive message, if there is no reply for a certain period of time and there is no reply even if the keep alive message is retransmitted, it is assumed that the content node has left the chord ring. .

(3) Step S-13
When the removal of another content node is detected, the operation differs depending on the direction of the separated content node on the chord ring. When the detached content node is downstream from the content node, the process proceeds to step S-A11. On the other hand, when the detached content node is upstream from the content node, the process proceeds to step S-B1.

(4) Step S-A11
A content search message with a ring repair option for searching for a detached content node from a downstream node as viewed from the detached content node is waited for a certain period of time.

(5) Step S-A12
When a content search message with a ring repair option for searching for a content node that has left the downstream node when viewed from the content node that has left is received within a predetermined time, the process proceeds to step S-A4. On the other hand, when that is not right, it transfers to step S-A13.

(6) Step S-A13
By referring to the finger table database, connection is attempted in order from the content node having an ID number that is larger than the ID number of the self-content node and closest. If the connection is successful, the process proceeds to step S-C11. If the connection could not be established, the finger table database is referred again, and the above operation is repeated until a connectable content node is found, and then the process proceeds to step S-C11.

(7) Step S-A14
When there is an efficient content search message transmission destination by referring to the finger table database as in the case of a normal content search message, the content search message is transmitted to the content node, and the process proceeds to step S-A11. On the other hand, if there is no efficient content search message destination by referring to the finger table database, that is, if the content node is the best recipient for content search messages with multiple ring repair options, the content search message sender The ID numbers of the content nodes are compared, and the connection is attempted in order from the content node having the closest ID number that is larger than the ID number of the own content node. All other content search messages are discarded.

(8) Step S-A15
A keep alive message is transmitted to the content node determined in step S-A14.

(9) Step S-B11
A content search message with a ring repair option for searching the detached content node is transmitted.

(10) Step S-B12
If a keep alive message is received within a certain time, the process proceeds to step S-C11. Otherwise, the process proceeds to step S-B3.

(11) Step S-B13
Resend a content search message with a ring repair option to search for a detached content node.

(12) Step S-B4
If a keep alive message is received within a certain time, the process proceeds to step S-C1. Otherwise, the process proceeds to step S-B5.

(13) Step S-B15
By referring to the finger table database, connection is attempted in order from the content node having the closest ID number that is smaller than the ID number of the self-content node. If the connection is successful, the process proceeds to step S-C11. If the connection could not be established, the finger table database is referred again, and the above operation is repeated until a connectable content node is found, and then the process proceeds to step S-C11.

(14) Step S-C1
Send and receive keep alive messages and repair the chord ring. The process proceeds to step S-1.

  Regarding the operation of the present embodiment, in step S-B11 to step S-B14, the content search message with the ring repair option for searching for the detached content node is retransmitted only once. Is not limited to once, and retransmission may be performed a plurality of times depending on the network state.

(Effect of the second embodiment)
Next, the effect of this embodiment will be described.

  According to the present embodiment, the code (Chord) ring has a content-oriented configuration, and the content node participates in the code (Chord) ring, thereby eliminating the need to register content location information in each node, Further, when a certain node leaves and the content node leaves the code (Chord) ring, the content node on the downstream side of the detached content node searches for the upstream content node of the detached content node, and connects the two to the code (Chord). ) By repairing the ring, the ring can be repaired when an arbitrary node leaves without giving the node redundant information.

  Also, according to the present embodiment, for example, in P2P-SIP (Session Initiation Protocol), IP (Internet Protocol) is obtained with the result obtained by applying SIP-URI (Uniform Resource Identifier) as content and SIP-URI as hash function as content ID. ) Registered in the code (Chord) ring together with the address, and when communicating, by searching for the IP address on the code (Chord) ring using the SIP-URI as a key, the terminals can communicate with each other as in the case of making a phone call. Also in the case of connecting, the effect of the present embodiment can be achieved.

(Third embodiment)
Next, a third embodiment of the present invention will be described in detail with reference to the drawings.

  Referring to FIG. 9, the present embodiment is similar to the first embodiment of the present invention in that the transmission / reception unit 1, the content node function unit 2, the content node generation function unit 3, and the content node deletion function unit. 4, a content node search function unit 5, a content database 6, and a finger table database 7.

  Furthermore, the present embodiment has a communication system program 8.

  The distributed authentication system program 8 includes a transmission / reception unit 1, a content node function unit 2, a content node generation function unit 3, a content node deletion function unit 4, a content node search function unit 5, a content database 6, and a finger The table database 7, the transmission / reception unit 1, the content node function unit 2, the content node generation function unit 3, the content node deletion function unit 4, the content node search function unit 5, the content database 6, and the finger The operation of the table database 7 is controlled. The transmission / reception unit 1, the content node function unit 2, the content node generation function unit 3, the content node deletion function unit 4, the content node search function unit 5, the content database 6, and the finger table database 7 are used for a communication system. Under the control of the program 8, the same processing as that performed by the communication system in the first embodiment is executed.

  Here, the hardware configuration of the node 100 will be described.

  FIG. 10 is a block diagram illustrating a hardware configuration example of the node 100 of the communication system according to the present embodiment.

  Referring to FIG. 10, a node 100 according to the present invention can be realized by a hardware configuration similar to that of a general computer device, and is a main memory such as a CPU (Central Processing Unit) 1001 and a RAM (Random Access Memory). A main storage unit 1002 used for a data work area and a temporary data save area; a communication control unit 1003 that transmits and receives data via the Internet 2000; a display unit 1004 such as a liquid crystal display, a printer and a speaker; a keyboard and a mouse The hard disk is composed of an input unit 1005 such as an interface unit 1006 that is connected to a peripheral device to transmit and receive data, a ROM (Read Only Memory), a magnetic disk, and a nonvolatile memory such as a semiconductor memory. Auxiliary storage unit 1007 is a disk device, and a system bus 1008 for connecting the above components of the information processing apparatus to each other.

  In the node 100 according to the present invention, its operation is implemented by mounting a circuit component composed of a hardware component such as an LSI (Large Scale Integration) in which a program for realizing such a function is incorporated in the node 100. As a matter of course, it can be realized in software by executing a program for providing each function of each component described above by the CPU 1001 on the computer processing apparatus.

  That is, the CPU 1001 loads the distributed authentication system program 8 stored in the auxiliary storage unit 1007 to the main storage unit 1002 and executes it, and controls the operation of the node 100 to control each function described above as software. Realize.

(Effect of the third embodiment)
Next, the effect of this embodiment will be described.

  According to the present embodiment, in order to obtain the same effect as the effects of the first and second embodiments of the present invention, the computer is communicated by using the distributed authentication system program 8 which is a communication system program. It can be operated as a system.

  In Example 1, an example of operation of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 11 is a flowchart illustrating an operation example of ring network restoration when the node 100 or the content node leaves the ring network, and FIG. 12 is a diagram illustrating state transition of the operation example.

  As a premise, the node ID 13, 16, 37, 42 (hereinafter referred to as node (13, 16, 37, 42)) indicates that it leaves the ring network (end message) to other nodes. It shall be withdrawn without notice.

  FIG. 13 shows the finger table with the nodes (13, 16, 37, 42) detached. In FIG. 13, the blank portion indicates the node ID of the node that has left, and the hatched portion indicates the location where the content search message transfer failed.

  Referring to FIGS. 11 to 13, the nodes (18, 48) whose upstream nodes are cut off are searched for the next node to the right and the node ID having the largest i entry value in its own finger table. A content search message (search message with error) with an error message (with IP address) is transmitted to the node (2) to perform a search (step S101).

  That is, the node (18) transmits the message to the node (52) to search for the node (16), and the node (48) does not respond even if the message is transmitted to the node (16). Therefore, the message is transmitted to the node (2) having the next largest i entry value to search for the node (42).

  Further, in step S201, the nodes (33, 8) whose downstream nodes are disconnected transition to the message waiting state.

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (52) receives the content search message from the node (18), the maximum node that does not exceed the search target node (16) is not itself on the finger table, so that the node (52) is the optimum recipient. Therefore, the content search message is transferred to the node (node (8)) having the largest node ID value that does not exceed the node (16). Upon receiving the content search message from the node (52), the node (8) forwards the content search message to the largest node (13) that does not exceed the search target node (16) on its finger table. Therefore, the largest node that does not exceed the search target node (16) is itself, and the node itself is the optimum recipient, so that the reception of the error message is waited.

  On the other hand, when the node (2) receives the content search message from the node (48), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (2) is the optimum receiver. However, since the transfer of the content search message to the largest node (37) that does not exceed the node (42) fails, the value of the i entry that does not exceed the node (42) becomes the next largest node (18). On the other hand, the content search message is transferred. When the node (18) receives the content search message from the node (2), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (18) is not the optimal recipient. However, if the content search message is transferred to the largest node (37) that does not exceed the node (42), it fails, so the value of the i entry that does not exceed the node (42) is the next largest node (27). Forward content search messages. When the node (27) receives the content search message from the node (18), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (27) is not the optimal recipient. However, if the content search message is transferred to the largest node (37) that does not exceed the node (42), it fails, so the value of the i entry that does not exceed the node (42) is the next largest node (33). Forward content search messages. Upon receiving the content search message from the node (27), the node (33) forwards the content search message to the largest node (37) not exceeding the search target node (42) on its finger table. Therefore, the largest node that does not exceed the node (42) is itself, and the node itself is the optimum recipient, so that it waits for reception of an error message.

  Next, when the error message reception standby state node (8, 33) does not receive the error message for a certain period of time, it reconnects with the source node of the content search message (step S103).

  That is, when no error message is received for a certain period of time, the ring network is restored by reconnecting the node (8) with the node (18) and reconnecting the node (33) with the node (48).

  In Example 2, one operation example of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 14 is a diagram illustrating state transition in the operation example.

  As a premise, in addition to the state of the first embodiment, it is assumed that the node (18) also leaves without notifying other nodes that the ring network is to be removed (end message).

  FIG. 15 shows a finger table in a state in which the nodes (13, 16, 18, 37, 42) are detached. In FIG. 15, the blank part indicates the node ID of the detached node, and the hatched part indicates the location where the transfer of the content search message has failed.

  Referring to FIG. 11, FIG. 14 and FIG. 15, the nodes (22, 48) whose upstream nodes are disconnected are searched for the right adjacent node that has left, and the value of the i entry is the highest on its finger table. A content search message (search message with error) to which an error message (with IP address) is added is transmitted to a node with a large node ID to perform a search (step S101).

  That is, the node (22) transmits the message to the node (56) to search for the node (18), and the node (48) does not respond even if the message is transmitted to the node (16). Therefore, the message is transmitted to the node (2) having the next largest i entry value to search for the node (42).

  Further, in step S201, the nodes (33, 8) whose downstream nodes are disconnected transition to the message waiting state.

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (56) receives the content search message from the node (22), the maximum node that does not exceed the search target node (16) is not itself on the finger table, so that the node (56) is the optimal recipient. Therefore, the content search message is transferred to the node (node (8)) having the largest node ID value that does not exceed the node (16). Upon receiving the content search message from the node (56), the node (8) forwards the content search message to the largest node (13) that does not exceed the search target node (16) on its finger table. Therefore, the largest node that does not exceed the search target node (16) is itself, and the node itself is the optimum recipient, so that the reception of the error message is waited.

  On the other hand, when the node (2) receives the content search message from the node (48), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (2) is the optimum receiver. However, if the content search message is transferred to the largest node (37) that does not exceed the node (42), it fails, and the value of the i entry that does not exceed the node (42) is the next largest node (18). If the content search message is transferred to the node (13) having the next largest i-entry value that does not exceed the node (42), the content search message also fails. The content search message is transferred to the node (8) having the next largest i entry value that does not exceed. When the node (8) receives the content search message from the node (2), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (8) is not the optimal recipient. Forwards the content search message to the largest node (27) not exceeding the node (42). When the node (27) receives the content search message from the node (8), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (27) is not the optimum receiver. However, if the content search message is transferred to the largest node (37) that does not exceed the node (42), it fails, so the value of the i entry that does not exceed the node (42) is the next largest node (33). Forward content search messages. When receiving the content search message from the node (27), the node (33) transfers the content search message to the largest node (37) not exceeding the search target node (42) on its finger table. However, since the largest node that does not exceed the node (42) is itself and becomes the optimum receiver, the state waits for the reception of the error message.

  Next, when the error message reception standby state node (8, 33) does not receive the error message for a certain period of time, it reconnects with the source node of the content search message (step S103).

  That is, when the error message is not received for a certain period of time, the node (8) reconnects with the node (22), and the node (33) reconnects with the node (48), thereby repairing the ring network.

  In Example 3, one operation example of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 16 is a diagram illustrating state transition in the operation example.

  As a premise, it is assumed that the nodes (27, 30, 37, 42) have left without notifying other nodes of the fact that they have left the ring network (end message).

  FIG. 17 shows the finger table with the nodes (27, 30, 37, 42) detached. In FIG. 17, the blank part indicates the node ID of the detached node, and the hatched part indicates the location where the transfer of the content search message has failed.

  Referring to FIG. 11, FIG. 16, and FIG. 17, each of the nodes (33, 48) whose upstream node is cut off searches for the right adjacent node that has left, and the value of the i entry is the highest on its finger table. A content search message (search message with error) to which an error message (with IP address) is added is transmitted to a node with a large node ID to perform a search (step S101).

  That is, the node (33) transmits the message to the node (2) to search for the node (30), and the node (48) transmits the message to the node (16) to transmit the node (42). Search for.

  In step S201, the nodes (22, 33) whose downstream nodes are cut off shift to the message waiting state.

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (2) receives the content search message from the node (33), the maximum node that does not exceed the search target node (30) is not itself on the finger table, so that the node (2) is the optimum recipient. Therefore, the content search message is transferred to the node (node (18)) having the largest i entry value that does not exceed the node (30). Upon receiving the content search message from the node (2), the node (18) forwards the content search message to the largest node (27) that does not exceed the search target node (30) on its finger table. Therefore, the content search message is transferred to the node (22) having the next largest i entry value that does not exceed the search target node (30). Upon receiving the content search message from the node (18), the node (22) forwards the content search message to the largest node (27) that does not exceed the search target node (30) on its finger table. Therefore, the largest node that does not exceed the node (30) is itself, and itself is the optimum recipient, so that reception of an error message is waited.

  On the other hand, when the node (16) receives the content search message from the node (48), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (16) is the optimum receiver. Therefore, the content search message is transferred to the largest node (33) that does not exceed the node (42). When receiving the content search message from the node (16), the node (33) transfers the content search message to the largest node (37) not exceeding the search target node (42) on its finger table. Therefore, the largest node that does not exceed the node (30) is itself, and itself is the optimum recipient, so that reception of an error message is waited.

  Next, when the error message reception standby node (22, 33) does not receive the error message for a certain period of time, it reconnects with the source node of the content search message (step S103).

  That is, when no error message is received for a certain time, the node (22) reconnects with the node (33), and the node (33) reconnects with the node (48), so that the ring network is restored.

  In Example 4, one operation example of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 18 is a diagram illustrating state transition in the operation example.

  It is assumed that the node (2, 3, 16, 18, 27, 30, 37, 42, 52) has left without notifying other nodes that it has left the ring network (end message). To do.

  FIG. 19 shows the finger table in a state in which the nodes (2, 3, 16, 18, 27, 30, 37, 42, 52) are detached. In FIG. 19, the blank portion indicates the node ID of the detached node, the shaded portion indicates the location where the content search message transfer failed, and the bold underline indicates the location where the finger table has been updated.

  11, 18, and 19, the nodes (8, 22, 33, and 48) whose upstream nodes are disconnected are searched for the right adjacent nodes that have left, and i entries are stored in their finger tables. A content search message (search message with error) to which an error message (with IP address) is added is transmitted to the node having the largest node ID, and the search is performed (step S101).

  That is, since the node (8) fails to transfer the content search message to the nodes (42, 27, 16) that do not exceed the search target node (3) on its finger table, The node (3) is searched by transmitting to the node (13) having the next largest i entry value. The node (22) transmits the message to the node (56) and searches for the node (18). In addition, since the node (33) fails to transfer the content search message to the nodes (2, 52, 42, 37) that do not exceed the search target node (30) in its own finger table, the search is failed. Since the largest node that does not exceed the target node (30) is itself and is the optimum recipient, the reception of the error message is awaited. In addition, since the node (48) fails to transfer the content search message to the nodes (16, 2) that do not exceed the search target node (42) on its finger table, the node (48) is referred to as i entry. Is sent to the next largest node (56) to search for the node (42).

  In step S201, the nodes (13, 22, 33, 61) whose downstream nodes are cut off shift to the message waiting state.

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (13) receives the content search message from the node (8), the maximum node that does not exceed the search target node (3) is not itself on the finger table, so that the node (13) is the optimum receiver. Therefore, the content search message is transferred to the node (node (48)) having the largest node ID value that does not exceed node (3). When the node (48) receives the content search message from the node (13), the maximum node that does not exceed the search target node (3) is not itself because it is not the optimum recipient on its finger table. However, even if the content search message is transferred to the largest node (2) that does not exceed the node (3), it fails, so the value of the i entry that does not exceed the node (3) becomes the next largest node (56). On the other hand, the content search message is transferred. When the content search message is received from the node (48), the node (56) is not the optimum recipient because the maximum node that does not exceed the search target node (3) is not itself on the finger table. However, even if the content search message is transferred to the largest node (2) that does not exceed the node (3), it fails, so the value of the i entry that does not exceed the node (3) becomes the next largest node (61). On the other hand, the content search message is transferred. Upon receiving the content search message from the node (56), the node (61) forwards the content search message to the largest node (2) not exceeding the search target node (3) on its finger table. Therefore, the largest node that does not exceed the node (3) is itself, and the node itself is the optimum recipient, so that it waits for reception of the error message.

  Further, when the node (56) receives the content search message from the node (22), the maximum node that does not exceed the search target node (18) is not itself on the finger table, so that the node (56) is the optimum recipient. Therefore, the content search message is transferred to the node (node (8)) having the largest i entry value that does not exceed the node (18). When the node (8) receives the content search message from the node (56), the maximum node that does not exceed the search target node (18) is not itself on the finger table, so that the node (8) is not the optimal recipient. However, even if the content search message is transferred to the largest node (16) that does not exceed the node (18), it fails. On the other hand, the content search message is transferred. Upon receiving the content search message from the node (8), the node (13) forwards the content search message to the largest node (16) that does not exceed the search target node (18) on its finger table. Therefore, the largest node that does not exceed the node (18) is itself, and the node itself is the optimum recipient, so that it waits for the reception of the error message.

  Furthermore, when the node (56) receives the content search message from the node (22), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (56) is the optimum recipient. However, even if the content search message is transferred to the largest node (27) that does not exceed the node (42), it fails, so the value of the i entry that does not exceed the node (42) is the next largest node ID. The content search message is transferred to the node (node (8)). When the node (8) receives the content search message from the node (56), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (8) is not the optimal recipient. However, even if the content search message is transferred to the largest node (27) that does not exceed the node (42), it fails, and the value of the i entry that does not exceed the node (42) is the next largest node (16). Therefore, the content search message is transferred to the node (13) having the next largest i entry value that does not exceed the node (42). When receiving the content search message from the node (8), the node (13) forwards the content search message to the largest node (30) not exceeding the search target node (42) on its finger table. Therefore, the content search message is transferred to the node (22) having the next largest i entry value that does not exceed the node (42). When the content search message is received from the node (13), the node (22) is not the optimum recipient because the maximum node that does not exceed the search target node (42) is not itself on the finger table. However, even if the content search message is transferred to the largest node (30) that does not exceed the node (42), it fails, and the value of the i entry that does not exceed the node (42) is the next largest node (27). Thus, even if the content search message is transferred, it fails, so that the largest node that does not exceed the node (42) is itself and is the optimum recipient, so that it waits to receive an error message.

  Next, when the error message reception standby state node (61, 13, 33, 22) does not receive the error message for a certain period of time, it reconnects with the source node of the content search message (step S103).

  That is, if no error message is received for a certain time, the node (61) reconnects with the node (8), the node (13) reconnects with the node (22), and the node (22) reconnects with the node (48). As a result, the ring network is restored, but the node (33) goes down because no content search message is transmitted from any node.

  For this reason, the communication network of Example 4 reconfigures a ring network by the nodes (8, 13, 22, 48, 56, 61).

  The connection between the node (48) and the node (56) is maintained by a normal sequence with one node (52) dropped (see FIG. 6).

  In Example 5, one operation example of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 20 is a diagram illustrating state transition in the operation example.

  As a premise, the node (3, 8, 13, 18, 27, 30, 37, 42, 52, 56, 61) does not notify other nodes that it has left the ring network (end message). It is assumed that it has left.

  FIG. 21 shows the finger table in a state where the nodes (3, 8, 13, 18, 27, 30, 37, 42, 52, 56, 61) are detached. In FIG. 21, the blank portion indicates the node ID of the detached node, the shaded portion indicates the location where the content search message transfer failed, and the bold underline indicates the location where the finger table has been updated.

  Referring to FIG. 11, FIG. 20, and FIG. 21, the nodes (2, 16, 33, and 48) whose upstream nodes are cut off are searched for the right adjacent node that has left, and i entries are entered in its own finger table. A content search message (search message with error) to which an error message (with IP address) is added is transmitted to the node having the largest node ID, and the search is performed (step S101).

  That is, the node (16) does not exceed the node (13) because the node (16) is not the optimum recipient because the maximum node that does not exceed the search target node (13) is not itself on the finger table. The message is transmitted to the node having the largest entry ID (node (48)) and the node (13) is searched. Also, the node (33) does not exceed the node (30) because the node (33) is not the optimum receiver because the largest node that does not exceed the search target node (30) is not itself on the finger table. The message is transmitted to the node having the largest entry ID (node (2)) to search for the node (30). The node (48) does not exceed the node (42) because the node (48) is not the optimum receiver because the maximum node that does not exceed the search target node (42) is not itself on the finger table. The message is transmitted to the node having the largest entry ID (node (16)) and the node (42) is searched. Further, node (2) fails even if it forwards the content search message to nodes (37, 18, 13, 8, 3) that do not exceed the search target node (61) on its finger table. Since the largest node that does not exceed the search target node (61) is itself, and the node itself is the optimum recipient, the reception of the error message is awaited.

  In step S201, the nodes (2, 22, 33, 48) whose downstream nodes are cut off shift to the message waiting state.

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (48) receives the content search message from the node (16), the maximum node that does not exceed the search target node (13) is not itself on the finger table, so that the node (48) is the optimum receiver. Therefore, the content search message is transferred to the node (node (2)) having the largest i entry value that does not exceed the node (13). Upon receiving the content search message from the node (48), the node (2) forwards the content search message to the largest node (8) not exceeding the search target node (13) on its finger table. Fails, even if the content search message is transferred to the node (3) having the next largest i-entry value that does not exceed the node (13), the maximum node that does not exceed the node (13) is unsuccessful. Since there is itself an optimum recipient, it waits for the receipt of an error message.

  Further, when the node (16) receives the content search message from the node (48), the maximum node that does not exceed the search target node (42) is not itself on the finger table, so that the node (16) is the optimum receiver. Therefore, the content search message is transferred to the node (node (33)) having the largest node ID value that does not exceed the node (42). When receiving the content search message from the node (16), the node (33) transfers the content search message to the largest node (37) not exceeding the search target node (42) on its finger table. And the largest node that does not exceed the node (42) is itself and is the optimum recipient, and thus waits to receive an error message.

  Next, when the error message reception standby state (2, 33) does not receive the error message for a certain period of time, it reconnects with the node that sent the content search message (step S103).

  That is, when no error message is received for a certain time, the node (2) reconnects with the node (16), and the node (33) reconnects with the node (48), so that the ring network is restored.

  Here, the optimum recipient node (2) receives the content search message with the node (16) as the transmission source and the content search message with the node (33) as the transmission source, but the value of the node ID is more In order to prioritize the content search message originating from the small node (16) and discarding the content search message originating from the node (33), the node (16) is reconnected.

  The connection between the node (16) and the node (22) is maintained by a normal sequence with one node (18) dropped (see FIG. 6).

  In Example 6, one operation example of ring network restoration according to the first embodiment will be described in more detail.

  FIG. 22 is a diagram illustrating state transition in the operation example.

  As a premise, compared with Example 5, the nodes (3, 8, 13, 18, 37, 42) are not originally joined on the ring network, and the nodes (27, 30, 52, 56, 61) are It is assumed that the node has left without notifying other nodes that the ring network is to be disconnected (end message).

  FIG. 23 shows a finger table in a state where nodes (3, 8, 13, 18, 37, 42) are not originally joined on the ring network and nodes (27, 30, 52, 56, 61) have left. Show. In FIG. 23, the dot portion indicates the node ID of a node that does not originally participate in the ring network, the blank portion indicates the node ID of the node that has left, and the shaded portion indicates the location where the content search message transfer failed. The bold underlined portion indicates the location where the finger table is updated.

  Referring to FIG. 11, FIG. 22, and FIG. 23, each of the nodes (2, 33) whose upstream node is cut off searches for the right adjacent node that has left, and the value of the i entry is the highest on its finger table. A content search message (search message with error) to which an error message (with IP address) is added is transmitted to a node with a large node ID to perform a search (step S101).

  That is, since the maximum node that does not exceed the search target node (61) is not itself, the node (2) does not exceed the node (61) because the maximum node that does not exceed the search target node (61) is not itself. The message is transmitted to the node having the largest entry value (node (16) = S node) and the node (61) is searched. Also, the node (33) does not exceed the node (30) because the node (33) is not the optimum receiver because the largest node that does not exceed the search target node (30) is not itself on the finger table. The message is transmitted to the node having the largest entry ID (node (2)) to search for the node (30).

  Next, the node that has received the message transfers the message according to a normal search sequence (step S102).

  That is, when the node (16) receives the content search message from the node (2), the maximum node that does not exceed the search target node (61) is not itself on the finger table, so that the node (16) is the optimum receiver. Therefore, the content search message is transferred to the node (node (48)) having the largest i entry value that does not exceed the node (61). Upon receiving the content search message from the node (16), the node (48) forwards the content search message to the largest node (52) not exceeding the search target node (61) on its finger table. And the largest node that does not exceed the node (61) is itself and is the optimum recipient, so that it waits for reception of an error message.

  In addition, when the node (2) receives the content search message from the node (33), the maximum node that does not exceed the search target node (30) is not itself on the finger table. Therefore, the message is transmitted to the node having the largest node ID (node (16) = S node) that does not exceed the node (30). When receiving the content search message from the node (2), the node (16) is not the optimum recipient because the maximum node that does not exceed the search target node (30) is not itself on the finger table. Therefore, the content search message is transferred to the node (node (22)) having the largest node ID value that does not exceed the node (30). When the node (22) receives the content search message from the node (16), the maximum node that does not exceed the search target node (30) is itself, and the node (22) is the optimum receiver. Then, it waits for reception of an error message.

  Next, when the node (48, 22) in the standby state for receiving the error message does not receive the error message for a certain period of time, it reconnects with the source node of the content search message (step S103).

  That is, when the error message is not received for a certain time, the node (48) reconnects with the node (2), and the node (22) reconnects with the node (33), so that the ring network is restored.

  In Example 7, one operation example when adding a node to the ring network according to the first embodiment will be described in detail.

  FIG. 24 is a diagram illustrating an initial configuration of a ring network according to the seventh embodiment.

  Referring to FIG. 24, as a premise, nodes (3, 10, 14, 29, 33, 48) are participating in the ring network.

  FIG. 25 is a diagram collectively showing the finger table database 7 included in each node according to the seventh embodiment.

  That is, referring to FIG. 25, for each node, node IDs and IP addresses of nodes before and after each node, and node IDs and IP addresses for each i entry are shown.

  Here, the operation when the node ID of the bootstrap node is the node (3) and the node (40) is added to the ring network will be described.

  FIG. 26 is a sequence diagram illustrating an operation when a node is added according to the seventh embodiment, and FIGS. 27 and 28 are diagrams illustrating an outline when a node is added.

  Step S201 FINDEROUTE: The node ID 40 notifies the node ID 3 of the node ID and the IP address of the node ID 40.

  Step S202: The node ID 3 responds to the node ID 40.

  Step S203 TESTRANGE: The node ID 3 requests the node ID 29 for the information of the S node of the node ID 40 based on the destination determined based on the finger table of the node ID 3.

  Step S204 / NOTINRANGE: The node ID 29 notifies the node ID 3 of the IP address of the node ID 33 determined based on the finger table of the node ID 29.

  Step S205 TESTRANGE: The node ID 3 requests the node ID 33 for information on the S node of the node ID 40.

  Step S206 · INRANGE: The node ID 33 notifies the node ID 3 of the IP address of the node ID 48 determined based on the finger table of the node ID 33.

  Step S207: The node ID 3 notifies the node ID 40 of the IP address of the node ID 48.

  Step S208: The node ID 40 responds to the node ID 3.

  Step S209 / NOTIFY: The node ID 40 notifies the node ID 48 of the IP address of the node ID 40.

  Step S210: The node ID 48 notifies the node ID 40 of the data of the node ID 48.

  Step S211: NOTIFY: The node ID 40 notifies the node ID 33 of the IP address of the node ID 40.

  Step S212: The node ID 33 notifies the node ID 40 of the data of the node ID 33.

  As described above, the node ID 40 newly participating in the ring network is notified of all the data of the adjacent nodes. Only adjacent nodes are guaranteed.

  When a node (40) newly joins, a node that has already joined may break a link with an adjacent node, but the network can be repaired by the above operation.

  FIG. 29 shows a finger table of the node ID 40, the node ID 33 adjacent to the node ID 40, and the node ID 48 after the above operation ends. The part in bold underline is the updated part.

  FIG. 30 shows more detailed data of the node ID 40, the node ID 33 adjacent to the node ID 40, and the node ID 48 after the above operation ends.

  As shown in FIG. 30, each node has node information of P node and S node in addition to its own node information.

  Although the present invention has been described with reference to the preferred embodiments, the present invention is not necessarily limited to the above embodiments, and various modifications can be made within the scope of the technical idea. .

It is the schematic which shows the physical structure of the communication system by the 1st Embodiment of this invention. It is the schematic which shows the logical structure of the communication system by the form of 1st Embodiment. It is a block diagram which shows the structure of the node by 1st Embodiment. It is the figure which showed collectively the finger table database 7 which each node by. It is a sequence diagram which shows the stabilization operation | movement by 1st Embodiment. It is a sequence diagram which shows the stabilization operation | movement when the node or content node by 1st Embodiment leaves | separates from a ring network. It is a flowchart which shows the more detailed operation example of ring network repair when the node 100 or content node by 1st Embodiment detaches | leaves from a ring network. It is a flowchart which shows operation | movement of the communication system by the 2nd Embodiment of this invention. It is a block diagram which shows the structure of the node by the 3rd Embodiment of this invention. It is a block diagram which shows the hardware structural example of the node 100 of the communication system by 3rd Embodiment. It is a flowchart which shows the operation example of ring network repair when the node 100 of Example 1 of this invention or a content node detaches | leaves from a ring network. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 1. FIG. It is a figure which shows the finger table of the state from which the node (13, 16, 37, 42) of Example 1 was detached. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 2 of this invention. It is a figure which shows the finger table of the state which the node (13,16,18,37,42) of Example 2 removed. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 3 of this invention. It is a figure which shows the finger table of the state from which the node (27, 30, 37, 42) of Example 3 removed. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 4 of this invention. It is a figure which shows the finger table of the state from which the node (2, 3, 16, 18, 27, 30, 37, 42, 52) of Example 4 left | separated. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 5 of this invention. It is a figure which shows the finger table of the state which the node (3, 8, 13, 18, 27, 30, 37, 42, 52, 56, 61) of Example 5 removed. It is a figure which shows the state transition of the operation example of the ring network restoration of Example 6 of this invention. A finger table in which the nodes (3, 8, 13, 18, 37, 42) originally did not participate in the ring network of the sixth embodiment and the nodes (27, 30, 52, 56, 61) have left FIG. It is a figure which shows the initial stage structure of the ring network of Example 7 of this invention. It is the figure which showed collectively the finger table database 7 which each node by Example 7 has. FIG. 20 is a sequence diagram illustrating an operation when a node is added according to the seventh embodiment. It is a figure explaining the outline at the time of adding the node by Example 7. FIG. It is a figure explaining the outline at the time of adding the node by Example 7. FIG. It is a figure which shows the finger table of node ID40 after node addition operation completion by Example 7, node ID33 adjacent to node ID40, and node ID48. It is a figure which shows the more detailed data of node ID40 after node addition operation completion by Example 7, node ID33 adjacent to node ID40, and node ID48.

Explanation of symbols

100: Node 1: Transmission / reception unit 1
2: Content node function unit 2
21: DHT participation function unit 22: DHT search function unit 23: DHT management unit 3: Content node generation function unit 3
4: Content node deletion function unit 4
5: Content node search function unit 5
6: Content database 6
7: Finger table database 7
8: Communication system program 8
1001: CPU
1002: Main storage unit 1003: Communication control unit 1004: Presentation unit 1005: Input unit 1006: Interface unit 1007: Auxiliary storage unit 1008: System bus 2000: Internet

Claims (40)

A P2P type network ring communication system composed of a plurality of nodes,
The node is
A content node function providing means generated for each one or a plurality of contents held by the own node ;
A communication system, wherein the content node function providing means participates in a participation type network ring formed by each content node function providing means . The content node function providing means for the content node function providing means detects that it has detached from the network ring adjacent,
Further located next to breakaway the content node function providing unit searches other content nodes function providing means for configuring the network ring in the content-oriented,
The communication system according to claim 1, wherein the network ring is repaired by executing a communication connection with the searched content node function providing means . When a plurality of the content node function providing means located adjacent to each other are separated from the network ring,
The content node function providing means for the content node function providing means adjacent detects that it has detached from the network ring,
Further located next to a plurality of content nodes function providing means for continuous withdrawal and said next searches other content nodes function providing means for configuring the network ring in the content-oriented,
3. The communication system according to claim 1, wherein the network ring is restored by executing a communication connection with the retrieved content node function providing unit . The node is
Transmitting / receiving means for transmitting / receiving information to / from other nodes connected by the network or the content node function providing means ;
A content database holding one or more content;
A content node that transmits to the content node function providing unit a content search request that requests a search for the content that the content node function providing unit is in charge of based on the unique identification information given to the content node function providing unit communication system according to any one of claims 1 to 3, characterized in Rukoto a search means. The node is
A finger table that stores the unique identification information in association with the IP address of the node that holds the content corresponding to the content node function providing means to which the unique identification information is assigned;
The content node search means includes:
The communication system according to claim 4, wherein the content search request is generated based on the unique identification information and the IP address of the finger table. The node is
The communication system according to claim 4 or 5, wherein when there are a plurality of finger tables in the own node, the finger tables are associated with each other and stored. The content node function providing means includes:
The communication system according to claim 5 or 6, wherein information of the finger table is exchanged with the other content node function providing means, and the finger table is updated.   The communication system according to claim 5 or 6, further comprising means for deleting the content node function providing means in charge of the deleted content when the content is deleted from the content database. The content node function providing means includes:
It generates a DHT join message for participating on the previous SL network,
The transmitting / receiving means includes
The communication system according to any one of claims 4 to 8, wherein the DHT participation message, the content search message, and the keep-alive message are transmitted to a transmission destination indicated by the message. . The transmitting / receiving means includes
Other of the node or other of the DHT join message from the content node function providing unit, a content retrieval message or keep-alive message, to send to the other the content node function providing means shown in the message The communication system according to claim 9, wherein the communication system is characterized. The content node function providing means includes:
One is generated for each content registered in the content database,
DHT participation function providing means for providing a function of registering the content node function providing means on the network ring;
DHT search means for transmitting the destination of the DHT participation message, the content search message, and the keep-alive message determined according to a predetermined algorithm described in the content search message;
11. The communication system according to claim 9, further comprising: a DHT management unit having a function of maintaining a network ring by transmitting and receiving keep-alive messages to / from another content node function providing unit . The content node function providing means includes:
When the DHT participation message is received from another content node function providing unit , the DHT participation function providing unit is caused to function,
12. The DHT search unit is caused to function when the content search request is received from the content node search unit and when the content search message is received from another content node function providing unit. The communication system according to 1. The DHT participation function providing means is:
12. The function of registering the content node function providing unit of the content in charge on the network ring by transmitting the DHT participation message to the other content node function providing unit. Or the communication system of Claim 12. The DHT search means includes:
Upon receiving the content node search request from the content node search means , after referring to the finger table related to the other content node function providing means existing in the same node other than the finger table related to the self content node function providing means , The communication system according to any one of claims 11 to 13, wherein a transmission destination of the message described in the content search message is determined according to a predetermined algorithm. The DHT search means includes:
When the content search message from the other content node function providing means is received, the finger table related to other content node function providing means existing in the same node other than the finger table related to the self content node function providing means is also referred to. The communication system according to claim 11, wherein a destination of the message described in the content search message is determined according to a predetermined algorithm. DHT management means
And other of the content node function providing means, mutually send and receive keepalive messages containing information of the finger table about its own content node function providing means together, and updates the information of the finger table, and maintains the ring The communication system according to any one of claims 11 to 15. DHT management means
When sending and receiving the keep alive message periodically, after sending the keep alive message, there is no reply for a certain period of time, and there is no reply even if the keep alive message is resent,
Communication system according to claim 16, by determining that the content node function providing means has left the network ring, and detecting a disengagement of the other of said content nodes function providing unit. DHT management means
When it is detected that a certain content node function providing means has left, and the detached content node function providing means is downstream as viewed from its own content node function providing means ,
Sent further from the downstream node when viewed from the breakaway the content node function providing unit, a network ring repair options marked with contents search message for searching the content node function providing unit that has left, characterized in that waiting a predetermined time The communication system according to claim 17. DHT management means
If there is the content node function providing means is detected that has left, and when leaving the said content node function providing means is an upstream when viewed from the own content node function providing means,
By transmitting a content search message with a network ring repair option for searching for the content node function providing means that has left the network using the DHT search means, the content node function providing means that exists further upstream as viewed from the content node function providing means that has left 19. The communication system according to claim 17, wherein the network ring is repaired by communicating with a content node function providing unit . DHT management means
When waiting for a content search message with a network ring repair option to search for the content node function providing means that has left,
Completing the repair of the network ring by transmitting / receiving keep-alive messages to / from the content node function providing means that has transmitted the content search message with the network ring repair option for searching the content node function providing means that has left the network. The communication system according to claim 18. DHT management means
Waiting for a content search message with a ring repair option to search for the content node function providing means that has left the network, and after the content search message with the network ring repair option to search for the content node function providing means has left a certain time If not,
Refer to the finger table database, try to connect in order from the content node function providing means having the identification information that is larger than and closest to the identification information given to the own content node function providing means ,
If the connection is successful, the content node function providing means and the network ring are repaired. If the connection is not successful, the finger table database is referred again, and the above operation is performed until the connectable content node function providing means is found. The communication system according to claim 18, wherein the communication system is repeated. DHT management means
In the case of waiting for the ring repair option Paste content search message for searching the content node function providing unit that has left, with the network ring repair option to search the content node function providing means in which a plurality of withdrawal during the predetermined time elapses Received a content search message,
When there is an efficient transmission destination of the content search message by referring to the finger table database as in the normal content search message, the content search message is transmitted to the content node function providing means . The communication system according to claim 18. DHT management means
If the content node function providing means is the optimum recipient for a plurality of content search messages with the network ring repair option referring to the finger table database,
Compares the value of the identification information given to the content node function providing means that is the transmission source of the content search message, and has the identification information that is larger than and closest to the identification information of the content node function providing means 23. The communication system according to claim 22, wherein connection is tried in order from the content node function providing means . DHT management means
Among the plurality of received content search messages, the identification information given to the content node function providing unit that is the transmission source of the content search message is larger and closest to the identification information of the own content node function providing unit 23. The communication system according to claim 22, wherein when the information is not identification information, all the content search messages are discarded. DHT management means
After sending a content search message with a network ring repair option to search for the content node function providing means that has left,
Claim 19 or claim 22, characterized in that waiting for a keepalive message from the content node function providing means waiting with the content retrieval message network ring repair option to search breakaway the content node function providing means The communication system according to 1. DHT management means
After sending a content search message with a network ring repair option to search for the content node function providing means that has left,
If you do not receive a keepalive message after a certain period of time,
The communication system according to any one of claims 23 to 25, wherein a content search message with a network ring repair option for searching for the content node function providing means that has left is retransmitted. DHT management means
If no keep-alive message is received for all the content search messages including retransmissions,
Refer to the finger table database, try to connect in order from the content node function providing means having the identification information that is smaller than and closest to the identification information given to the own content node function providing means ,
If the connection is successful, the content node function providing means and the network ring are repaired. If the connection is not successful, the finger table database is referred again, and the above operation is performed until the connectable content node function providing means is found. 27. The communication system of claim 26, wherein the communication system repeats.   The communication system according to any one of claims 1 to 27, wherein the P2P type network ring has a configuration of a code ring. A communication method in a communication system of a P2P type network ring composed of a plurality of nodes ,
The content node function providing means included in the node participates in a participation type network ring formed by each content node function providing means ;
In the content node function providing means adjacent detects that it has detached from the network ring said content node function providing means,
Further located next to breakaway the content node function providing means, retrieving other content nodes function providing means for configuring the network ring in the content-oriented,
Have a step of repairing the network ring by performing a communication connection with the found the content node function providing means,
The content node function providing means is generated for each one or a plurality of contents held by the node . When a plurality of the content node function providing means located adjacent to each other are separated from the network ring,
In the content node function providing means adjacent detects that it has detached from the network ring said content node function providing means,
Searching for another content node function providing means that configures the network ring to be content-oriented , which is located further next to the plurality of adjacent continuous content node function providing means that have left the network;
30. The communication method according to claim 29, further comprising a step of repairing the network ring by executing a communication connection with the retrieved content node function providing unit . In the DHT management means provided in the content node function providing means,
Regularly send and receive keep-alive messages with other content node function providing means, and after sending keep-alive messages, if there is no reply for a certain period of time and there is no reply even if the keep-alive message is resent, the other Detecting that the content node function providing means has left the network ring;
Confirming the direction on the network ring of the other content node function providing means that has left, with reference to its own content node function providing means ;
When the direction of the separated other content node function providing means is downstream with respect to its own content node function providing means ,
A waiting step for waiting for reception of a breakaway the other content nodes function network ring further searching the other content nodes function providing means separated from the downstream node from the providing means repair options marked with contents search message,
If not received, the unique identification information given to the content node function providing means, and the IP address of the node that holds the content corresponding to the unique identification information content node function providing means has been granted Other than the unique identification information that is larger than the unique identification information given to the content node function providing means and closest to the unique identification information. Trying to connect in order from the content node function providing means ;
If received, refer to the finger table database to check whether there is an efficient destination of the content search message; and
31. The communication method according to claim 29, further comprising a step of transmitting a keep-alive message to the confirmed content node function providing means of the efficient transmission destination.   32. The communication method according to claim 31, wherein in the waiting step, the content search message is waited for a predetermined time. In the DHT management means provided in the content node function providing means,
Regularly send and receive keep-alive messages with other content node function providing means, and after sending keep-alive messages, if there is no reply for a certain period of time and there is no reply even if the keep-alive message is resent, the other Detecting that the content node function providing means has left the network ring;
Confirming the direction on the network ring of the other content node function providing means that has left, with reference to its own content node function providing means ;
If the direction of withdrawal and said other content nodes function providing means upstream with respect to the own content node function providing means,
Sending a content search message with a network ring repair option to search for the other content node function providing means that has left;
An acknowledgment step for confirming whether or not a keep-alive message has been received;
33. The communication method according to claim 29, further comprising a step of returning a keep-alive message when a keep-alive message is received. In the DHT management means,
When the keep alive message is not received, refer to the finger table database that stores the unique identification information and the IP address of the node that holds the content to which the unique identification information is stored in association with each other, Attempting to connect in order from the other content node function providing means having the unique identification information that is smaller than the unique identification information given to the own content node function providing means and closest to the unique identification information; 34. The communication method according to claim 33, comprising:   The communication method according to claim 33 or 34, wherein in the reception confirmation step, reception of a keep-alive message is waited for a predetermined time. A communication program executed on a node constituting a P2P type network ring,
In the content node function providing means provided in the node ,
Realize the function to participate in the participation type network ring formed by each content node function providing means ,
To the content node function providing means on the node that detects that the content node function providing means adjacent is disconnected from the network ring,
Further located next to breakaway the content node function providing unit, a function to search other content nodes function providing means for configuring the network ring in the content-oriented,
Realizing a function of repairing the network ring by executing a communication connection with the searched content node function providing means ,
The communication program, wherein the content node function providing means is generated for each of one or a plurality of contents held by the node . When a plurality of the content node function providing means located adjacent to each other are separated from the network ring,
To the content node function providing means on the node that detects that the content node function providing means adjacent is disconnected from the network ring,
A function of searching for another content node function providing unit that configures the network ring in a content-oriented manner , which is located next to the plurality of adjacent continuous content node function providing units that have left;
37. The communication program according to claim 36, wherein a function for restoring the network ring is realized by executing a communication connection with the retrieved content node function providing means . In the DHT management means provided in the content node function providing means,
Regularly send and receive keep-alive messages with other content node function providing means, and after sending keep-alive messages, if there is no reply for a certain period of time and there is no reply even if the keep-alive message is resent, the other A function of detecting that the content node function providing means has left the network ring;
A function of confirming the direction of the other content node function providing means that has left on the network ring with reference to its own content node function providing means ;
When the direction of the separated other content node function providing means is downstream with respect to its own content node function providing means ,
Waiting feature to wait for reception of a breakaway the other content nodes function network ring further searching the other content nodes function providing means separated from the downstream node from the providing means repair options marked with contents search message a predetermined time When,
If not received within a predetermined time, the holding of the content corresponding to the content node capabilities and unique identification information given to providing means, a content node function providing unit granted the unique identification information The unique identification that is larger than the unique identification information given to the content node function providing means and is closest to the unique identification information with reference to the finger table database that stores the IP address of the node in association with each other A function of trying to connect in order from the other content node function providing means having information;
A function that, when received within a predetermined time, refers to the finger table database and confirms whether there is an efficient destination of the content search message;
38. The communication program according to claim 36 or 37, wherein the confirmed content node function providing means of the transmission destination realizes a function of transmitting a keep alive message. In the DHT management means provided in the content node function providing means,
Regularly send and receive keep-alive messages with other content node function providing means, and after sending keep-alive messages, if there is no reply for a certain period of time and there is no reply even if the keep-alive message is resent, the other A function of detecting that the content node function providing means has left the network ring;
A function of confirming the direction of the other content node function providing means that has left on the network ring with reference to its own content node function providing means ;
If the direction of withdrawal and said other content nodes function providing means upstream with respect to the own content node function providing means,
A function of transmitting a content search message with a network ring repair option for searching for the other content node function providing means that has left;
A reception confirmation function for confirming whether or not the keep-alive message is received within a predetermined time;
The communication program according to any one of claims 36 to 38, wherein a function of returning a keep-alive message when a keep-alive message is received within a predetermined time is realized. In the DHT management means,
A finger table that stores the unique identification information and the IP address of the node that holds the content to which the unique identification information is associated in a case where a keep-alive message is not received within a predetermined time. Refer to the database, and in order from the other content node function providing means having the unique identification information that is smaller than the unique identification information given to the own content node function providing means and closest to the unique identification information 40. The communication program according to claim 39, wherein a function for attempting connection is realized.

Images