JP5233799B2 - Content distribution system, node device, content distribution method, and content acquisition processing program - Google Patents

Description

  The present invention relates to a technical field of a peer-to-peer (P2P) type communication system including a plurality of node devices that can communicate with each other via a network.

  With regard to this type of system, Patent Document 1 discloses a technology relating to peer-to-peer type content distribution. Specifically, in this technique, each node device searches for another node device that stores the content, and requests the content from any one of the searched node devices. Thereby, each node apparatus acquires a content.

JP 2006-197400 A

  By the way, there is a case where it is desired to distribute the same content to a plurality of node devices. In this case, if all the node devices request the content all at once, the load is concentrated on the node device storing the content. Therefore, when distributing content, for example, the number of node devices that request (acquire) content is initially small, and the number of node devices that request content increases with the passage of time. Conceivable. The node device that first acquired the content can be distributed to other node devices first. Therefore, the more node devices that have acquired the content, the more node devices that can be distributed.

  However, the line speed may differ depending on the type of line (for example, FTTH (Fiber To The Home), ADSL (Asymmetric Digital Subscriber Line), etc.) to which each node device is connected. In this case, when a node device with a low line speed goes to acquire content for the first time, it takes more time than necessary until the content is acquired and can be distributed to other node devices. Therefore, there is a problem that it takes a long time until the number of node devices that can distribute content increases.

  Accordingly, the present invention has been made in view of the above points and the like, and a content distribution system capable of shortening the time required for content distribution when node devices having different line speeds coexist. An object is to provide a node device, a content distribution method, and a content acquisition processing program.

In order to solve the above-described problem, the invention according to claim 1 is a content distribution system including a plurality of node devices capable of communicating with each other via a network, wherein the node device is configured so that each of the plurality of node devices includes the node device. Order information indicating that the plurality of node devices are acquired in order from the group with the highest line speed, as the acquisition order of the contents of the plurality of groups in which the plurality of node devices are grouped according to the line speed of the line for connecting to the network ; Group information acquisition means for acquiring time information for each group for determining whether to acquire content, and time information indicating a shorter time interval as the line speed is faster; and the group information acquisition means based on the obtained sequence and temporal information, the group information obtainment hand among the plurality of groups A determination means for probabilistically determining whether to acquire content for each time interval of the belonging group among the time intervals indicated by the time information at a time according to the belonging group to which the node device belongs. when it is determined that the acquisition by means and a content obtaining means for obtaining a content, that Ru comprising a.

  According to the present invention, acquisition of content by a node device to which the node belongs belongs in order from the group with the highest line speed. Therefore, the number of node devices that can transmit content to other node devices increases rapidly. Thereby, the time required for content distribution can be shortened.

According to a second aspect of the invention, a node device in a content distribution system having a plurality of node devices mutually communicable through a network, the line for connection to said plurality of node devices each said network As the content acquisition order of a plurality of groups in which the plurality of node devices are grouped according to the line speed, it is determined whether to acquire the order information indicating that the line speed is acquired in order, and the content. Group information acquisition means for acquiring time information indicating a time interval for each group, the time information indicating a shorter time interval as the line speed is faster, and the order information and time information acquired by the group information acquisition means A node device comprising the group information acquisition means of the plurality of groups A determination unit that probabilistically determines whether or not to acquire content for each time interval of the group belonging to the time interval indicated by the time information at a time according to the belonging group. And a content acquisition unit that searches for a node device that can transmit the content as a result of acquiring the content and acquires the content from the searched node device.

  According to the present invention, when each node device acquires content, the node device searches for a node device that can transmit the content because the content has been acquired. Therefore, each node device can acquire content from among many node devices capable of transmitting content at an early stage from the start of content distribution. Thereby, the time required for content distribution can be shortened.

According to a third aspect of the present invention, in the node device according to the second aspect, the group information acquisition means uses the transmission rate information indicating the number of transmittable contents of each node device for each group, and the line speed is high. Further acquiring transmission number information in which the number of possible transmissions is set to a larger value for a group, and the determination unit is configured to determine each group based on the order information and the transmission number information acquired by the group information acquisition unit. Content with a probability corresponding to the number of node devices that can transmit the content and the number of transmittable devices and the number of node devices that do not acquire the content among the node devices that belong to the belonging group. It is characterized by determining that it is acquired .

  According to this invention, the acquisition of content by the node device to which the node belongs is started in the order of the group having the largest number of transmissions. Then, the number of node devices corresponding to the number of node devices that can transmit the content and the number of transmittable groups belonging to the node device each acquire the content from the node device that can transmit the content. . As a result, the rate of increase in the number of node devices capable of transmitting content can be increased.

According to a fourth aspect of the present invention, in the node device according to the second or third aspect, whether or not it is time to perform the determination by the determination unit based on the order information acquired by the group information acquisition unit. A timing determination unit that further includes a timing determination unit that accelerates the timing of the determination by the determination unit as the acquisition order of the belonging group is earlier, and the determination unit performs the determination by the determination unit If it is determined by the time determination means, whether to acquire content is determined probabilistically .

  According to this invention, in order from the group with the highest line speed, the node device belonging to the group acquires the content, and then the node device belonging to the next group starts acquiring the content. Therefore, as the predetermined time interval for acquiring the contents of each group, the line speed of the group may be considered. Therefore, the time required for content distribution can be shortened by adjusting the predetermined time interval of each group.

According to a fifth aspect of the present invention, there is provided a computer for a node device in a content distribution system including a plurality of node devices that can communicate with each other via a network. As the content acquisition order of a plurality of groups in which the plurality of node devices are grouped according to the line speed, it is determined whether to acquire the order information indicating that the line speed is acquired in order, and the content. A group information acquisition step for acquiring time information indicating a time interval for each group, the time information indicating a shorter time interval as the line speed is faster, and the order information and time information acquired by the group information acquisition step The group information acquisition step of the plurality of groups is executed based on A determination step that probabilistically determines whether content is acquired for each time interval of the belonging group among the time intervals indicated by the time information at a time according to the group to which the node device belongs; and the determination step when it is determined that the acquired result, searches the node devices enabled sending the content by acquired the content, the content acquisition step of acquiring the content from the retrieved node apparatus, thereby to execute It is characterized by.

According to a sixth aspect of the present invention, in the content distribution method in the content distribution system including a plurality of node devices that can communicate with each other via a network, the node device connects to the network of each of the plurality of node devices. As the acquisition order of the contents of a plurality of groups in which the plurality of node devices are grouped according to the line speed of the communication line, order information indicating that acquisition is performed in order from the group with the highest line speed is acquired. A group information acquisition step for acquiring time information indicating a time interval for each group, the time information indicating a time interval indicating a shorter time interval as the line speed is higher, and the node device, the group information acquisition step based on the obtained sequence and temporal information, the serial the grayed among the groups Probably determine whether to acquire content for each time interval of the belonging group among the time intervals indicated by the time information at a time corresponding to the group to which the node device that executed the group information acquisition step belongs. a determination step, the node device, when the it is determined to be obtained by determination step, characterized in that it comprises a content acquisition step of acquiring the content and.

  According to the present invention, the number of node devices that can transmit to other node devices increases quickly. Thereby, the time required for content distribution can be shortened.

It is a figure which shows an example of the connection aspect of each node apparatus in the content distribution storage system S which concerns on one Embodiment. It is a figure which shows an example of a mode that a content is delivered in the content distribution storage system S which concerns on one Embodiment. It is a figure which shows the calculation method of the number of winning and a winning probability in case a some group has already acquired the content. It is a figure which shows an example of the content set to the parameter information which concerns on one Embodiment. It is a figure which shows an example of the acquisition time of the content of each group which concerns on one Embodiment in time series. It is a figure showing an example of outline composition of center server SA concerning one embodiment. It is a figure showing an example of outline composition of node Nn concerning one embodiment. It is a flowchart which shows the process example at the time of the line speed measurement of the content distribution storage system S which concerns on one Embodiment. It is a flowchart which shows the process example at the time of parameter information preparation of the content distribution storage system S which concerns on one Embodiment. It is a flowchart which shows the process example at the time of the content delivery of the content distributed storage system S which concerns on one Embodiment. It is a flowchart which shows the process example in the content acquisition process of the control part 21 of the node Nn which concerns on one Embodiment.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, the best embodiment of the invention will be described with reference to the drawings. The embodiment described below is an embodiment when the present invention is applied to a content distributed storage system.

[1. Outline and configuration of distributed content storage system]
First, with reference to FIG. 1 etc., the structure and operation | movement outline | summary of the content distribution preservation | save system which concern on this embodiment are demonstrated.

  FIG. 1 is a diagram showing an example of a connection mode of each node device in the content distributed storage system S according to the present embodiment.

  As shown in the lower frame 101 of FIG. 1, IX (Internet eXchange) 3, ISP (Internet Service Provider) 4a, 4b, DSL (Digital Subscriber Line) line operators (devices) 5a, 5b, FTTH (Fiber To A network 8 such as the Internet is constructed by the line provider (device) 6 and the communication line 7. The network 8 is a real-world communication network. Note that a router for transferring data (packets) is appropriately inserted in the network (communication network) 8 in the example of FIG. 1, but the illustration is omitted in FIG. For example, a telephone line or an optical cable is used as the communication line 7.

  A plurality of node devices (hereinafter referred to as “nodes”) Nn (n = 1, 2, 3,...) Are connected to such a network 8. Each node Nn is assigned a unique manufacturing number and an IP (Internet Protocol) address. The content distributed storage system S according to the present embodiment is a peer-to-peer network system formed by participation of any of a plurality of nodes Nn, as shown in the upper frame 100 of FIG. It has become.

  The network 9 shown in the upper frame 100 of FIG. 1 is an overlay network 9 that configures a virtual link formed using the existing network 8. The overlay network 9 is a logical network. The overlay network 9 is realized by a specific algorithm, for example, an algorithm using DHT. Each node Nn participating in the content distributed storage system S (in other words, the overlay network 9) is assigned a node ID, which is unique identification information having a predetermined number of digits.

  The node ID is, for example, a value obtained by hashing an IP address or manufacturing number individually assigned to each node Nn using a common hash function. The node IDs are arranged in a distributed manner in the ID space. For example, SHA-1 or the like is used as the hash function. The hashed value has a bit length of 160 bits, for example. The node IDs are distributed in the ID space without being distributed.

  The participation in the distributed content storage system S is performed by a non-participating node Nn (for example, the node N8) transmitting a participation message indicating a participation request to any participating node Nn. . The arbitrary node Nn is, for example, a contact node that always participates in the system S.

  Each node Nn holds a routing table using DHT. This routing table defines the transfer destinations of various messages on the content distributed storage system S. Specifically, a plurality of pieces of node information including node IDs, IP addresses, and port numbers of nodes Nn that are moderately separated in the ID space are registered in this routing table.

  One node participating in the content distributed storage system S stores the minimum necessary node information of the node Nn as a routing table. By transferring various messages between the nodes Nn, node information about the node Nn not knowing (not storing) the node information is acquired.

  Such a routing table using DHT is well known in Japanese Patent Application Laid-Open No. 2006-197400 and the like, and will not be described in detail.

  The content distribution storage system S stores (stores) various content replicas having different contents in a predetermined file format in a distributed manner in a plurality of nodes Nn. The node Nn storing the replica is hereinafter referred to as a “content holding node”. The content distribution storage system S can use a replica between the nodes Nn. The original of each content is stored in the center server SA. For example, a replica of the content of a movie with the title XXX is stored in the node N5. On the other hand, a replica of the movie content whose title is YYY is stored in the node N3. In this way, it is distributed and stored in a plurality of nodes Nn (hereinafter referred to as “content holding nodes”).

  Information such as a content name (title) and a content ID (identification information unique to each content) is added to each of the above-described content replicas. This content ID is generated, for example, by hashing the content name + an arbitrary numerical value with a hash function common to the case of obtaining the node ID. Alternatively, the system administrator may give a unique ID value for each content. In this case, the content catalog information in which the correspondence between the content name and the content ID is written is distributed to all the nodes Nn. In this case, the content ID has the same bit length as the node ID.

  The locations of the content replicas that are distributed and stored are stored as index information by a node Nn (hereinafter referred to as “root node”) that manages (stores) the location of the content replicas. The index information includes a set of node information of the node Nn that stores the replica, a content ID of the content, and the like. Such a root node is determined to be, for example, the node Nn having the node ID closest to the content ID (for example, the higher-order digits match more).

  For example, index information about a replica of a movie content whose title is XXX is managed by a node N4 which is the root node of the content (content ID). Further, for example, the index information about the content replica of the movie whose title is YYY is managed by the node N7 which is the root node of the content (content ID). Further, such a root node is determined to be a node Nn having a node ID closest to the content ID, for example. The node ID closest to the content ID is, for example, a node ID that matches the content ID more frequently with the upper digits.

  When a user of a certain node Nn wants to acquire a replica of the desired content, the node Nn that desires to acquire the replica (hereinafter referred to as “user node”) generates a message (query). This message (query) is a content location inquiry message including the content ID of the content desired to be acquired and the IP address of the user node. The content location inquiry message is also a message for searching for a content holding node. The above-described content location inquiry message is sent to another node Nn according to the DHT routing table acquired by the user node. That is, the user node sends a content location inquiry message toward the root node. As a result, the content location inquiry message finally arrives at the root node by DHT routing using the content ID as a key.

  In each node Nn, the attribute information such as the content name and content ID of the content is described in, for example, content catalog information distributed from the center server SA to all the nodes Nn.

  The content ID included in the content location inquiry (search) message may be generated by the user node by hashing the content name with the common hash function. The DHT routing is known in Japanese Patent Application Laid-Open No. 2006-197400 and the like, and thus detailed description thereof is omitted.

  The root node that has received the content location inquiry message acquires (searches) index information corresponding to the content ID included therein from the index cache. The acquired index information is returned to the user node that is the transmission source of the content location inquiry message. The user node that has acquired the index information in this way can download (acquire) a replica of the content based on the index information. The user node accesses the content holding node based on the IP address of the content holding node included in the index information. It becomes possible to download (acquire) a replica of the content from the accessed content holding node. Note that the index information may include, for example, node information of a plurality of content holding nodes (when replicas of the same content are stored in a plurality of content holding nodes). In such a case, the user node selects one content holding node from among the plurality of content holding nodes. The user node can then connect to the selected content holding node and download a replica of the content.

  The root node transmits a content transmission request message to the content holding node indicated by the IP address or the like included in the index information, whereby the user node downloads the replica from the content holding node. You can also. The user node can also acquire the index information from a cache node that caches the same index information as the root node until the content location inquiry message reaches the root node.

  Also, when the user node acquires and stores a content replica from the content holding node, the stored user node generates a publish (registration notification) message. The publish message is a message for informing the root node that the replica has been saved. The publish message includes the content ID of the replica and its own node information. The publish message is sent toward the root node (addressed to the root node). As a result, the publish message arrives at the root node by DHT routing using the content ID as a key, like the content location inquiry message. Then, the root node receives the publish message. The root node registers (stores in the index cache area) index information including a set of node information and content ID included in the publish message. In this way, the user node becomes a new content holding node that holds a replica of the content.

[2. Overview of content delivery]
Next, content distribution in the content distributed storage system S will be described with reference to FIGS. In the following description, the replica stored in each node Nn is also referred to as content.

  In the distributed content storage system S, a plurality of nodes Nn participating in the system are grouped into a plurality of groups according to the transmission speed of the line connected to the network 8, that is, the line speed. In this system, in this embodiment, for example, content acquisition is started in order from the group with the highest line speed. First, how content is distributed will be described.

  FIG. 2 is a diagram illustrating an example of a state in which content is distributed in the content distributed storage system S according to the present embodiment.

  For example, it is assumed that the nine nodes N1 to N9 shown in FIG. 2 belong to the group 1 with the fastest line speed. In addition, only the center server SA initially holds content to be distributed to all nodes Nn. With the start of content distribution, for example, one node Nn among the nodes N1 to N9 acquires content.

  Of the nodes Nn belonging to the group 1, which node Nn acquires the content at which timing is not determined in advance. This is determined by probability calculation. In the example illustrated in FIG. 2, one node Nn out of nine acquires content. Accordingly, each of the nodes N1 to N9 executes a lottery process for winning content acquisition with a probability of 1/9. Only the selected node Nn, for example, the node N1, acquires the content. In order to acquire content, the node N1 inquires the location of the content to the root node (acquires index information by transmitting a content location inquiry message to the root node). At this point, since there is no node Nn that stores the content to be acquired, the node N1 downloads the content from the center server SA.

  After a predetermined time has elapsed since the start of content acquisition by the node N1, the next two nodes Nn acquire content. Here, since the two nodes Nn of the eight nodes N2 to N9 that have not yet acquired the content acquire the content, the winning probability is 2/8. Each of the node nodes N2 to N9 performs a lottery process with a winning probability of 2/8. Then, for example, the nodes N2 and N3 that have been elected start acquiring content. When the nodes N2 and N3 query the root node for the location of the content, the node N1 can be identified as the content holding node. Therefore, the nodes N2 and N3 download contents from the node N1, respectively. The node N1 uploads contents to two nodes Nn at the same time.

  After a predetermined time has elapsed from the start of content acquisition at the nodes N2 and N3, next, six nodes Nn acquire content. Here, since the six nodes Nn among the six nodes N4 to N9 that have not yet acquired the content acquire the content, the winning probability is 1. Accordingly, all of the nodes N4 to N9 start to acquire content. When the nodes N4 to N9 query the root node for the location of the content, it is possible to specify that the nodes N1 to N3 are content holding nodes. Therefore, the nodes N4 to N9 download the content from any one of the nodes Nn among the nodes N1 to N3. At this time, the nodes N4 to N9, for example, randomly select a node Nn as a content acquisition destination from the nodes N1 to N3. Therefore, each of the nodes N1 to N3 uploads content to the two nodes Nn at the same time on average. In this way, content is distributed.

  In the example shown in FIG. 2, one node Nn uploads content to two nodes Nn at the same time. This number (hereinafter referred to as “uploadable number”) is set in advance, for example, according to the uplink bandwidth of the line speed on the uploading side. The higher the line speed, the higher the number of nodes that can be uploaded at the same time. In the example shown in FIG. 2, the nodes N1 to N9 belong to the same group 1. That is, the line speeds of the nodes N1 to N9 are the same or fall within a predetermined range. Accordingly, the number of nodes N1 to N9 that can be uploaded is the same (two units), and this number is the number of groups 1 that can be uploaded.

  Further, the time from when the node N1 starts to acquire the content until the nodes N2 and N3 start to acquire the content needs to be equal to or longer than the time required for the node N1 to download the content. Similarly, the time from when the nodes N2 and N3 start acquiring the content until the nodes N4 to N7 start acquiring the content needs to be equal to or longer than the time required for the nodes N2 and N3 to download the content. Specifically, this time (hereinafter referred to as “standby time”) is set in advance according to the upstream bandwidth of the line speed on the content uploading side and the downstream bandwidth of the line speed on the content downloading side. In the example shown in FIG. 2, since the nodes N1 to N9 belong to the same group 1, the standby times are the same. Further, since there is no group having a line speed higher than that of group 1, the waiting time may be determined by considering the line speed of group 1.

  Furthermore, the number of nodes Nn that acquire content each time the standby time elapses (hereinafter referred to as “winning number”) is determined by the number of nodes Nn that have already acquired content and the number of uploadable devices. In the example shown in FIG. 2, after the node N1 acquires content, the number of nodes Nn that have already acquired content is one. Therefore, the next winning number will be 1 × uploadable number 2 = 2. The next winning number will be 3 x number of uploadable units 2 = 6. Then, the next winning number is 9 × uploadable number 2 = 18. Note that the number of nodes Nn that initially acquire content is not limited to one, but may be a plurality.

  In this way, each node Nn belonging to the group acquires content at any one of the timings (acquisition timings) that arrive each time the standby time elapses.

  In determining the number of winning units, if content has already been acquired in a plurality of groups, the number of uploadable units in each group is considered.

  FIG. 3 is a diagram illustrating a method of calculating the winning number and the winning probability when a plurality of groups have already acquired content.

  For example, as shown in FIG. 3, it is assumed that 20 nodes Nn belong to group 1, 60 nodes belong to group 2, and 300 nodes Nn belong to group 3. Here, the line speed increases in the order of group 1, group 2, and group 3. Further, it is assumed that the number of uploadable groups in group 1 is 3, and the number of uploadable groups in group 2 is 1.

  In group 1, all 20 nodes Nn have already acquired content. Also in group 2, all 30 nodes Nn have already acquired content. On the other hand, in Group 3, no content has been acquired yet.

  In this state, the 300 nodes Nn belonging to the group 3 calculate the winning number and the winning probability in the lottery process. In group 1, 20 nodes Nn perform uploading to 3 nodes Nn, each of which can be uploaded. Therefore, content is uploaded to 20 × 3 = 60 nodes Nn. In group 2, 60 nodes Nn upload to one node Nn, which is the number of uploadable nodes. Accordingly, the content is uploaded to 60 × 1 = 60 nodes Nn. The sum of the number of uploads in group 1 and the number of uploads in group 2 is 120. This is the winning number for Group 3. Since all the nodes Nn in group 3 have not yet acquired the content, the winning probability is 120/300 = 0.4.

  As described with reference to FIG. 2, each node Nn selects, for example, a node Nn as a content acquisition destination at random. Then, contents are requested from the same number of nodes Nn to each node Nn belonging to group 1 and to each node Nn belonging to group 2. Then, the node Nn belonging to the group 1 can afford to upload the content, but the node Nn belonging to the group 2 has a heavy load. Therefore, in the present embodiment, the actual number of uploaded contents is adjusted to the number of uploadable items in each group. Specifically, when the user node that acquires content acquires the index information of the content holding node from the root node, the user node selects a predetermined number of nodes Nn from among the nodes Nn indicated by the index information. Next, the user node accesses the plurality of selected nodes Nn, and downloads a part of the content from each node Nn. Downloading a part of the content may be performed, for example, by downloading for a certain period of time or by downloading a certain number of bytes. Each user node that has downloaded a part of the content calculates the respective reception speed. Then, the user node continues to download the content only from the node Nn having the fastest reception speed, and interrupts the download to the other nodes Nn. Then, the actual number of uploads is adjusted according to the line speed of the uploading node Nn. The final number of uploads converges to the number of uploadable groups in the group to which each node Nn belongs.

  FIG. 4 is a diagram illustrating an example of contents set in the parameter information according to the present embodiment.

  Each node Nn obtains parameter information (an example of order information and transmission number information) from the center server SA in order to perform the lottery process and to determine the timing for performing the lottery process. As shown in FIG. 3, the parameter information includes, for example, identification information indicating a group in that order in association with each sequence number from which content is acquired, the number of nodes Nn belonging to the group (hereinafter referred to as “number of units”). The waiting time and the number of uploadable devices are set. The order starts from 1 and G indicates the last order. Hereinafter, the I-th group is indicated as group I. This I is also referred to as a group number. The smaller the group number is, the faster the line speed becomes.

In addition, _{M I} is an affiliation number of group I. _TI is the waiting time of group I. The waiting time T _I is set based on the upstream bandwidth of the line speed of the groups T _{1 to} T _I and the downstream bandwidth of the line speed of the group T _I. The waiting time T _I, the node Nn in the group I is the value of more than the time required for downloading the content is set. Then, the waiting time T _I, the acquisition order of the content has become shorter as fast. In addition, _{U I} is the upload possible number of group I. Uploadable number U _I is set on the basis of the upstream bandwidth of the line speed of the Group I. For example, the upload allows the number U _I, number to the extent that no load on the line is set to be as much as possible. Then, you can upload the number U _I, the acquisition order of the content has become so much faster. Note that there are cases where the number of uploadable units is set to the same value between groups in which the acquisition order is adjacent.

  In the present embodiment, the ratio of the upstream bandwidth to the downstream bandwidth at the line speed of each node Nn can be adjusted. Depending on the line speed (group to which it belongs), each node Nn may upload content to a plurality of nodes Nn at the same time. On the other hand, each node Nn downloads content only from one node Nn at a time. Thus, for example, the line can be used efficiently by setting the upstream bandwidth of each node Nn wider than the downstream bandwidth.

In the parameter information, K is indicated as the group number of the group (own group) to which the node Nn that acquired the parameter belongs. Therefore, the number of units belonging to group K, the standby time, and the number of uploads are M _k , T _k, and U _k . Actually, the node ID indicates which node Nn belongs to each group instead of the group number K to which it belongs.

  Next, how content is distributed to each group will be described. Here, as shown in FIG. 4, for example, the number of groups 1 belonging to group 1, the standby time, and the number of uploadable units are 20 units, 60 seconds, and 3 units. In addition, the number of groups 2, the waiting time, and the number of uploadable units are 30 units, 180 seconds, and 1 unit.

  FIG. 5 is a diagram showing an example of content acquisition timing of each group in time series.

  At the start of content distribution, first, one node Nn belonging to group 1 starts content acquisition. Then, after 60 seconds, which is the standby time of the group 1, has elapsed since the start of content distribution, among the nodes Nn belonging to the group 1, the number of content acquired 1 × number of uploadable 3 = 3 nodes Nn acquire content. After 120 seconds from the start of content distribution, the number of nodes Nn whose distribution has been completed is four. Therefore, next, 12 nodes Nn among the nodes Nn belonging to the group 1 acquire contents. After 180 seconds from the start of content distribution, the number of nodes Nn that have been distributed is 16. Therefore, the next winning number is 48 units. However, among the nodes Nn belonging to the group 1, there are four nodes Nn that have not acquired content at this time. Accordingly, the four nodes Nn acquire content.

Here, t _I is the time required from the start of content distribution to the completion of content distribution to all the nodes Nn in the I-th group. After 240 seconds from the start of content distribution, content distribution is completed for all nodes Nn in group 1. The content acquisition period of group 1 is between 240 seconds from the start of content distribution. And t ₁ = 240. At this point, the node Nn belonging to the group 2 starts to acquire content. Here, the winning number is 60. Accordingly, all 30 nodes Nn belonging to the group 2 acquire content. When 180 seconds, which is the standby time of group 2, elapses and t ₂ = 420 seconds elapses from the start of content distribution, the content for group 1 belonging to group ₁ + group 2 belonging to group M ₂ = 50 Delivery is complete. The content acquisition period of group 2 is between 420 seconds and 240 seconds after the start of content distribution.

In this way, after all the nodes Nn belonging to a certain group have acquired the content, the node Nn belonging to the next group in the next order starts acquiring the content. For example, after all the nodes Nn belonging to the group K-1 have finished acquiring the content, the nodes Nn belonging to the group K start acquiring the content. After _{t k} seconds elapsed from the content delivery _start, the content delivery is completed for the _{M 1 + M 2 + ··· M} K table of the node Nn. Then, after elapse of t _G seconds from the start of content distribution, the content distribution is completed for M ₁ + M ₂ +... _MG units, that is, all nodes Nn.

  Thus, content acquisition is started from a group having a high line speed. In each group, during the content acquisition period, each node Nn to which it belongs acquires content at any one of the timings (acquisition times) that arrive each time the standby time elapses. Further, after all the nodes Nn belonging to the previous group have acquired the content, the node Nn belonging to the next group starts to acquire the content.

  Next, a description will be given of the difference in time required to complete the distribution when the content is distributed in this way and when the content is not distributed without being grouped.

As a first example, a case where the standby time is changed according to the line speed will be described. In this example, it is assumed that groups 1 and 2 exist. The line speed of group 1 is faster than that of group 2. The number of groups M ₁ , the waiting time T _1, and the uploadable number U ₁ of the group ₁ are 256 units, 60 seconds, and 1 unit. Further, the number M ₂ belonging to the group 2, the waiting time T ₂ and the uploadable number U ₂ are 256 units, 180 seconds, and one unit. In order to facilitate understanding, the number of units belonging to both groups and the number of uploadable units are set to the same value.

In this case, one node Nn belonging to the group 1 acquires content at the start of content distribution. The number of nodes Nn whose distribution is completed after 60 seconds has elapsed is one. Thereafter, every 60 seconds, the number of nodes Nn for which distribution is completed increases to 2, 4, 8, 16, 32, 64, 124, and 256. In this way, distribution is completed to all the nodes Nn belonging to the group 1 after 9 × T ₁ = 540 seconds from the start of content distribution. Next, all 256 nodes Nn belonging to group 2 acquire content. Then, distribution is completed for all nodes Nn 540 seconds + 1 × T ₂ = 720 seconds after the start of content distribution.

On the other hand, when the grouping is not performed, the node Nn that should belong to the group 1 and the node Nn that should belong to the group 2 are mixed. Therefore, it is necessary to adjust the standby time to 180 seconds of the group 2 having the slowest line speed. Further, the number of uploadable units is assumed to be an average value, for example. In this example, (U ₁ × M ₁ + U ₂ × M ₂ ) / (M ₁ + M ₂ ) = 1.

  Then, one node Nn acquires content at the start of content distribution. The number of nodes Nn whose distribution is completed after 180 seconds has elapsed is one. Thereafter, every 180 seconds, the number of nodes Nn whose distribution is completed increases to 2, 4, 8, 16, 32, 64, 124, 256, and 512. In this way, distribution is completed for all the nodes Nn after 10 × 180 seconds = 1800 seconds have elapsed from the start of content distribution.

  From the above results, it can be seen that the time required to complete the content distribution can be shortened by grouping and acquiring the content in order from the group with the highest line speed.

As a second example, a case where the number of uploadable devices is changed according to the line speed will be described. Also in this example, it is assumed that groups 1 and 2 exist. The line speed of group 1 is faster than that of group 2. The number of groups M ₁ , the waiting time T _1, and the uploadable number U ₁ of the group ₁ are 256 units, 60 seconds, and 3 units. Further, the number M ₂ belonging to the group 2, the waiting time T ₂ and the uploadable number U ₂ are 256 units, 60 seconds, and 1 unit.

  The faster the line speed, the greater the number of uploadable units. Therefore, the number of uploadable groups in group 1 is larger than that in group 2. Further, if the waiting time is the same, the more the number of uploadable units, the faster the content can be transmitted to many nodes Nn. Therefore, the faster the line speed, the shorter the time required for content distribution completion.

In this case, one node Nn belonging to the group 1 acquires content at the start of content distribution. The number of nodes Nn whose distribution is completed after 60 seconds has elapsed is one. Thereafter, every 60 seconds, the number of nodes Nn for which distribution is completed increases to 4, 16, 64, and 256. In this way, distribution is completed to all the nodes Nn belonging to the group 1 after 5 × T ₁ = 300 seconds from the start of content distribution. Next, all 256 nodes Nn belonging to group 2 acquire content. Then, the distribution is completed to all the nodes Nn after 300 seconds + 1 × T ₂ = 360 seconds from the start of content distribution.

On the other hand, when the grouping is not performed, the standby time is 60 seconds as before. In addition, the uploadable number is (U ₁ × M ₁ + U ₂ × M ₂ ) / (M ₁ + M ₂ ) = 2.

  Then, one node Nn acquires content at the start of content distribution. Then, the number of nodes Nn whose distribution is completed after 60 has elapsed is one. Thereafter, every 60 seconds, the number of nodes Nn for which distribution is completed increases to 3, 9, 27, 81, 243, and 256. In this way, distribution is completed to all the nodes Nn after 7 × 60 seconds = 420 seconds from the start of content distribution.

  From the above results, it can be seen that the time required to complete the content distribution can be shortened by grouping and acquiring the contents in order from the group with the highest line speed.

  In the first and second examples, only one of the waiting time and the uploadable number is changed for each group. It is more effective if both the waiting time and the number of uploadable units are changed.

  In the description so far, the case has been described where the content is distributed all at once, but the same can be applied to the case where the content is divided and distributed sequentially. For example, the content is divided into a plurality of data (hereinafter referred to as “fragment data”), and these fragment data are stored in the center server SA. Each piece of fragment data is assigned a content ID in the same way as normal content. Therefore, each node Nn can acquire each piece of fragment data in the same manner as normal content. Referring to FIG. 2 as an example, at the start of content distribution, the node N1 downloads the first fragment data among the plurality of fragment data constituting the content from the center server SA. After the standby time has elapsed, the node N1 downloads the second fragment data from the center server SA. At the same time, the nodes N2 and N3 download the first fragment data from the node N1, respectively. After the standby time has elapsed, the node N1 downloads the third fragment data from the center server SA. At the same time, the nodes N2 and N3 download the second fragment data from the node N1, respectively. At the same time, the nodes N4 to N9 download the first fragment data from any of the nodes N1 to N3, respectively. In this way, fragment data is distributed one after another. The waiting time at this time is longer than the time required for downloading the fragment data.

[3. Configuration of center server SA]
Next, the configuration and functions of the center server SA will be described with reference to FIG.

  FIG. 6 is a diagram illustrating a schematic configuration example of the center server SA.

  As shown in FIG. 6, the center server SA stores and saves a control unit 11 composed of a CPU having a calculation function, a working RAM, a ROM for storing various data and programs, and various data and various programs ( Storage unit 12 composed of an HD or the like for storing) and a communication unit 13 for controlling communication of information with the node Nn through the network 8 or the like. The unit 12 and the communication unit 13 are connected to each other via a bus 14.

  The storage unit 12 stores the node ID, IP address, port number, and the like of each node Nn. The storage unit 12 stores content to be distributed. The storage unit 12 stores line speed information for all nodes Nn in association with node IDs. The line speed information is information indicating the line speed of the corresponding node Nn. Further, the storage unit 12 stores parameter information for distributing content.

  The control unit 11 performs overall control of the entire center server SA when the CPU reads and executes a program stored in the storage unit 12 or the like.

  For example, the control unit 11 acquires the line speed of each node Nn from each node Nn as line speed information. Then, the control unit 11 creates parameter information. Specifically, the control unit 11 performs grouping of the nodes Nn based on the acquired line speed information. This grouping may be performed, for example, by determining the line speed range of each group in advance, and the control unit 11 determining which node Nn is in which line speed range. Further, for example, the control unit 11 may flexibly determine the line speed range for each group in accordance with the line speed distribution of each node Nn. Also, the number of groups may be determined flexibly. The standby time and the number of downloadable units for each group may be set in advance, for example. In addition, for example, the control unit 11 may flexibly determine the standby time and the downloadable number of each group according to the number of units belonging to each group, the line speed, and the like. When determining the number of units belonging to each group, the standby time, and the number of uploadable units, the control unit 11 creates parameter information based on these pieces of information.

  The control unit 11 that created the parameter information distributes the parameter information to all the nodes Nn. This distribution of parameter information may be performed, for example, for all nodes Nn participating in the content distributed storage system S by DHT multicast (multicast message transfer processing) disclosed in Japanese Patent Laid-Open No. 2007-053662. good.

  Further, when content is input to the center server SA from the outside, the control unit 11 transmits a distribution start command to all the nodes Nn participating in the content distributed storage system S. As a result, distribution of content is started. Note that, for example, the content may be stored in the storage unit 12 in advance, and the control unit 11 may transmit a distribution start command when the date and time for starting the distribution of content arrives.

  The above program may be downloaded from a predetermined server on the network 8, for example. The program may be recorded on a recording medium such as a CD-ROM and read via a drive of the recording medium.

[4. Configuration and function of node Nn]
Next, the configuration and function of the node Nn will be described with reference to FIG.

  FIG. 7 is a diagram illustrating a schematic configuration example of the node Nn.

  As shown in FIG. 7, each node Nn has a control unit 21 as a computer composed of a CPU having a calculation function, a working RAM, a ROM for storing various data and programs, and various data and various programs. A storage unit 22 composed of an HD (hard disk) or the like for storing (storing), a buffer memory 23 for temporarily storing a received content replica or the like, and encoded video data included in the content replica ( Video information) and audio data (audio information), etc., and decoder unit 24 for decoding (data decompression, decoding, etc.), and video processing for performing a predetermined drawing process on the decoded video data and outputting the video signal Unit 25 and C for displaying video based on the video signal output from the video processing unit 25 T, a display unit 26 such as a liquid crystal display, an audio processing unit 27 that converts the decoded audio data into an analog audio signal after D (Digital) / A (Analog) conversion, and amplifies this by an amplifier, and outputs the analog audio signal; A speaker 28 that outputs an audio signal output from the audio processing unit 27 as a sound wave, a communication unit 29 for controlling communication of information between other nodes Nn and the like through the network 8, and an instruction received from a user And an input unit (for example, a keyboard, a mouse, a remote controller, an operation panel, or the like) 30 that gives an instruction signal corresponding to the control unit 21 to the control unit 21. The decoder unit 24, the communication unit 29, and the input unit 30 are connected to each other via a bus 31. As the node Nn, a personal computer, an STB (Set Top Box), or the like can be applied in addition to a dedicated server device.

  The storage unit 22 includes a routing table using DHT, index information, and content catalog information, and an IP address and port number of a contact node that is an access destination when participating in the content distributed storage system S, and the center server SA. The IP address, port number, etc. are stored.

  The storage unit 22 stores parameter information acquired from the center server SA.

  The control unit 21 reads and executes a program stored in the storage unit 22 or the like (including the content acquisition processing program of the present invention) by the CPU, whereby the group information acquisition unit, the acquisition time determination unit, and the content acquisition in the present invention Functions as a means.

  The control unit 21 as the group information acquisition unit receives the parameter information distributed from the center server SA.

  Further, when receiving the distribution start command transmitted from the center server SA, the control unit 21 determines the content acquisition time of the own node based on the parameter information received in advance as the acquisition time determination means. Specifically, every time the standby time elapses, the control unit 21 determines whether it is time to perform a lottery process for the own group to which the own node belongs. This standby time is set as the standby time of the group that performs the lottery process at that time. Which group of waiting times is used is determined based on the parameter information. When the control unit 21 determines that it is time to perform the lottery process of the own group, the control unit 21 executes the lottery process. In this lottery process, the control unit 21 executes the content acquisition process when the content acquisition is won. On the other hand, when the content acquisition is not won, the control unit 21 executes the lottery process again after the standby time of the own group has elapsed. Detailed processing contents will be described later.

  The control unit 21 that has been selected to acquire the content acquires the content from the content holding node (the node Nn that can distribute the content) as a content acquisition unit. Here, the control unit 21 searches for a content holding node. Specifically, the control unit 21 transmits a content location inquiry message toward the root node. And the control part 21 receives the index information transmitted from the root node. Thereby, the content holding node is searched.

  Then, the control unit 21 downloads content from the searched content holding node. Here, as described above, the control unit 21 first downloads a part of content from a plurality of content holding nodes. Then, the control unit 21 continues to download content from the content holding node having the fastest reception speed among these content holding nodes.

  The content acquisition processing program may be downloaded from a predetermined server on the network 8, for example, or recorded on a recording medium such as a CD-ROM, for example, via the drive of the recording medium. It may be read.

[5. Operation of Content Distributed Storage System S]
Next, the operation of the distributed content storage system S according to the present embodiment will be described with reference to FIGS.

  FIG. 8 is a flowchart showing an example of processing at the time of line speed measurement of the distributed content storage system S according to the present embodiment.

  The process of the node Nn illustrated in FIG. 8 is executed periodically, for example. First, in step S1, the control unit 21 of the node Nn measures the line speed of the own node (step S1). For example, the control unit 21 transmits a predetermined amount of measurement data to a predetermined line speed measurement server. The server for line speed measurement measures the reception time at this time, and transmits this as reception time information to the requesting node Nn. The control unit 21 that has received the reception time information calculates the uplink bandwidth of the line speed based on the data transmission amount and the reception time. The line speed measurement server transmits a predetermined amount of measurement data to the requesting node Nn. The control unit 21 that has received the measurement data calculates the reception speed at this time and sets it as the downlink bandwidth of the line speed.

  Next, in step S2, the control unit 21 transmits the line speed information and the node ID of the own node to the center server SA (step S2). Here, the control unit 21 integrates the uplink bandwidth and the downlink bandwidth of the calculated line speed to obtain line speed information.

  The control unit 11 of the center server SA that has received the line speed information and the node ID stores the information in the storage unit 12 in association with each other in step S11 (step S11). In this way, the line speed information of each node Nn is collected by the center server SA.

  FIG. 9 is a flowchart showing an example of processing at the time of creating parameter information of the content distributed storage system S according to the present embodiment.

  The processing of the center server SA shown in FIG. 9 is executed periodically, for example. First, in step S21, the control unit 11 creates parameter information based on the line speed information of each node Nn (step S21). Specifically, the control unit 11 performs grouping based on the line speed of each node Nn indicated by the line speed information. Subsequently, the control part 11 calculates | requires the affiliation number of each group. Further, the control unit 11 determines the standby time and the number of downloadable units for each group. Then, the control unit 11 creates parameter information from these pieces of information and stores the parameter information in the storage unit 12.

  Next, in step S22, the control unit 11 transmits the created parameter information to each node Nn (step S22).

  The control unit 21 of the node Nn that has received the parameter information stores the parameter information in the storage unit 22 in step S31 (step S31). At this time, when old parameter information is stored in the storage unit 22, the control unit 21 overwrites the received new parameter information.

  FIG. 10 is a flowchart showing an example of processing at the time of content distribution of the content distributed storage system S according to the present embodiment.

  The processing of the center server SA shown in FIG. 10 is executed, for example, when content to be distributed is input to the center server SA. First, in step S41, the control unit 11 transmits a distribution start command to each node Nn (step S41).

  The control unit 21 of the node Nn that has received the distribution start command starts a timer in step S51 (step S51). Specifically, the control unit 21 sets 0 to the timer t. This timer t is a timer variable to which 1 is added every second. The value of this timer variable indicates the elapsed time from the start of content distribution.

Next, in step S52, the control unit 21 performs constant setting and initial value setting (step S52). Specifically, the control unit 21 determines from the parameter information stored in the storage unit 22 the group number K to which the node belongs, which is a constant, the number M _{1 to} M _{G of} each group, and the waiting time T _{1 to} T. _G, set the upload possible number _U 1 ~U _G. The control unit 21, as an initial value of a variable, group number I, set enforcement number i, the content delivery completion time _{t 0} and winning already number _H 1 to H _G. The group number I is a variable indicating which group currently performs the lottery process. Since the content distribution is started, the initial value of the group number I is 1. The number of executions i is a variable indicating the number of lottery processes for the group currently performing the lottery process. Since the first lottery process of group 1 is performed, the initial value of the enforcement count i is 0. Content distribution completion time t _I is, the time at which all the nodes Nn belonging to the Group I has finished acquiring the content, which is a variable that shows the elapsed time from the content delivery start. Here, the content distribution completion time of the group in the previous order is used as information indicating the time when the group currently performing the lottery process starts the lottery process. Since group 1 starts the lottery process, T _i-1 = T ₀ = 0 is set as an initial value. The winning numbers H _{1 to} H _G are variables indicating the number of nodes Nn that have been won for content acquisition in the lottery process for each group. Since no node Nn is elected at the start of content distribution, the initial values of the elected units H _{1 to} H _G are all 0.

Next, in step S53, the control unit 21 determines whether or not the timer t is equal to t _I-1 + T _I × i (step S53). In other words, the control unit 21 determines whether or not it is time for the group I to execute the lottery process. At this time, if the timer t is not equal to t _I-1 + T _I × i, that is, if it is not the timing at which the group I should execute the lottery process (step S53: NO), The process proceeds to S53 and the same determination is repeated.

On the other hand, in step S53, the control unit 21 determines that the timer t is equal to t _I-1 + T _I × i, that is, if it is time for the group I to execute the lottery process (step S53: YES). i = i + 1 is calculated (step S54). Next, in step S55, the control unit 21 determines whether or not the group number I is 1 and the enforcement count i is 1 (step S55). That is, the control unit 21 determines whether or not it is the first lottery process (at the start of content distribution) of the first group. At this time, when the group number I is 1 and the number of enforcements i is 1 (step S55: YES), the control unit 21 sets the winning number h to 1 (step S56). The winning number h is a variable indicating the number of items that are won for content acquisition in the current lottery process. Since content distribution is started, a constant is set as the winning number h.

On the other hand, in step S55, if the group number I is not 1 or the enforcement count i is not 1 (step S55: NO), the control unit 21 calculates the winning number h (step S57). If it is not at the start of content distribution, the node Nn that has already acquired the content by winning the acquisition of the content transmits the content to as many nodes Nn as the number of uploadable members in the group to which it belongs. The total number of nodes Nn to which the content is transmitted becomes the winning number h. Therefore, the winning number h is obtained by adding the values obtained by multiplying the number of nodes Nn already won to the uploadable number for each group from group 1 to group I. That means
h = U ₁ × H ₁ + U ₂ × H ₂ +... + U _I−1 × H _I−1 + U _I × H _I.
It becomes. However, for the groups from Group 1 to Group I-1, all nodes Nn have already won. Therefore, h is obtained by the following equation 2.
h = U ₁ × M ₁ + U ₂ × M ₂ +... + U _I−1 × M _I−1 + U _I × H _I.
It becomes.

When the process of step S56 or S57 is completed, the control unit 21 determines in step S58 whether the group number I is equal to the group number K of the own group (step S58). That is, the control unit 21 determines whether or not it is time for the own group to execute the lottery process. At this time, when the group number I is not equal to the group number K of the own group, that is, when the own group is not at the timing for executing the lottery process (step S58: NO), the control unit 21 has H _I = H _I + h is calculated (step S59). That is, the control unit 21 increases the number of groups that have been won in the group I by the number of units that have been won this time.

Then, in step S60, the control unit 21 determines whether the winning already number _{H I} is affiliation number _{N I} above (step S60). That is, the control unit 21 determines whether or not all the nodes Nn belonging to the group I have been won for content acquisition. At this time, the control unit 21, if elected already number H _I is less than the affiliation number N _I, that is, when the node Nn which is not elected to get the contents of the node Nn in the group I is present (Step S60: NO), the process proceeds to Step S53. That is, the control unit 21, since the lottery process group I is not finished, it is to wait the waiting time T _I.

On the other hand, in step S60, the control unit 21, if elected already number H _I is affiliation number N _I above, i.e., if all the nodes Nn in the group I was elected to obtain a content (step S60: _YES), calculates the _{t I = t I-1 +} T I × i ( step S61). That is, since all the lottery processes for group I have been completed, the control unit 21 sets a time for starting the lottery process for group I + 1. Next, in step S62, the control unit 21 calculates I = I + 1 and sets i to 0 (step S62). That is, the control unit 21 sets the target for the lottery process as group I + 1. After completing this process, the control unit 21 proceeds to step S53.

In step S57, when the group number I is equal to the group number K of the own group, that is, when the own group has a timing for executing the lottery process (step S57: YES), the control unit 21 performs steps S64 to S66. The lottery process is executed. First, the control unit 21 calculates a winning probability P for winning content acquisition (step S63). The winning number P is the ratio of the node Nn to be won this time among the nodes Nn of the own group that has not been won for content acquisition yet. Therefore, the winning number P is obtained by the following expression 3.
P = h / (N _K −H _k ) Equation 3

Next, in step S64, the control unit 21 determines whether or not the winning probability P is greater than 1 (step S64). At this time, when the winning probability P is 1 or less (step S64: NO), the control unit 11 generates a random number not less than 0 and less than 1, and sets the random number to the random value R (step S65). Next, in step S66, the control unit 21 determines whether or not the winning probability P is larger than the random value R (step S66). That is, the control unit 21 determines whether or not the content acquisition is won. At this time, if the winning probability P is equal to or less than the random value R, that is, if the winning probability P is lost (step S66: NO), the control unit 21 proceeds to step S59. That is, the control unit 21 increases the selected number H _k of the group K by the number won this time, waits for the waiting time T _k , and then performs the lottery process again.

  On the other hand, in step S66, when the winning probability P is larger than the random value R, that is, when the winning is won, the control unit 21 executes content acquisition processing described later (step S67). Moreover, also when the winning probability P is larger than 1 in step S64 (step S64: YES), the control unit 21 executes content acquisition processing (step S67). In some cases, the number h of winners may exceed the number of nodes Nn that have not yet been won for content acquisition. In this case, the winning probability P exceeds 1. At this time, since 100% is won for content acquisition, the control unit 21 omits unnecessary processing.

  FIG. 11 is a flowchart illustrating a processing example in the content acquisition processing of the control unit 21 of the node Nn according to the present embodiment.

  As shown in FIG. 11, in step S71, the control unit 21 transmits a content location inquiry message to the root node (step S71). Next, in step S72, the control unit 21 receives the index information transmitted from the root node (step S72). Next, in step S73, the control unit 21 determines the number of content holding nodes with reference to the received index information (step S73). At this time, when the number of content holding nodes is 0 (step S73: 0), the control unit 21 downloads content from the center server SA (step S74).

  In step S73, when the number of content holding nodes is 1 (step S73: 1), the control unit 21 downloads all of the content from the one content holding node (step S75).

  In step S73, when the number of content holding nodes is two or more (step S73: two or more), the control unit 21 downloads a part of the content from a predetermined number of content holding nodes (step S76). ). At this time, when the number of content holding nodes is less than the predetermined number, the control unit 21 downloads from all content holding nodes, for example. Next, in step S77, the control unit 21 calculates each reception speed for downloading a part of the content (step S77). Next, in step S78, the control unit 21 continues the download from the content holding node having the fastest reception speed, and downloads all the content (step S78). On the other hand, the control unit 21 interrupts downloading from other content holding nodes.

  When the download of the content is completed (when the process of step S74, S75, or S78 is completed), the control unit 21 transmits a publish message to the root node (step S79). As a result, the own node is registered as a content holding node. When this process is completed, the control unit 21 ends the content acquisition process. Moreover, the control part 21 will complete | finish the process shown in FIG. 10, if a content acquisition process is complete | finished.

  As described above, according to the present embodiment, the control unit 21 of the node Nn receives the parameter information set so that the content acquisition order is accelerated as the line speed increases. Moreover, the control part 21 determines whether it is a timing which a self group performs a lottery process based on the received parameter information. At this time, if it is time for the own group to perform the lottery process, the control unit 21 performs the lottery process at the standby time interval of the own group. And when the control part 21 wins by a lottery process, a content is acquired from a content holding node. In this way, the control unit 21 determines the content acquisition time. As a result, the content is acquired by each group by acquiring the content at any acquisition time of the acquisition times that each node Nn belonging to each group arrives at the waiting time interval. Then, content acquisition (lottery processing) by each group is started in order from the group with the earliest acquisition order indicated by the parameter information.

  Therefore, the number of nodes Nn that can transmit content to other nodes Nn increases quickly. Therefore, the time required for content distribution can be shortened.

  In addition, the control unit 21 acquires index information by transmitting a publish message toward the root node. In this way, the control unit 21 searches for a content holding node. Then, the control unit 21 acquires content from the searched content holding node.

  Therefore, each node Nn can acquire the content from the content holding nodes increased at an early stage from the start of content distribution. Therefore, the time required for content distribution can be shortened.

  In the parameter information received by the control unit 21, the number of uploadable units, which is larger as the line speed is higher, is set for each group. Then, the control unit 21 calculates the number of nodes h to be won for content acquisition by calculating Equation 2. That is, as the number of uploadable groups in the group to which the content holding node belongs increases, the number of nodes Nn to be elected increases.

  Therefore, the rate of increase in the number of content holding nodes, that is, the number of nodes Nn that can distribute content can be increased.

  In addition, after all the nodes Nn belonging to the previous group from the first group to the last group in order of acquisition according to the parameter information are elected to acquire the content, the control unit 21 uses the node Nn belonging to the next group. The timing of the lottery process of each group is specified so that the lottery process is performed. And the control part 21 determines the acquisition time of the content of a self-node by performing a lottery process at this timing.

  Accordingly, the waiting time for each group may be determined by considering the line speed of that group. Therefore, by adjusting the waiting time of each group, it is possible to reduce the time required for content distribution.

  In the above embodiment, the uploadable number is set for each group according to the line speed, but this may be a fixed value. That is, the uploadable number of all groups may be the same value. The number of uploadable devices in this case may be stored in advance by each node Nn (may be hard-coded in the content acquisition program) and may not be included in the parameter information.

  Further, in the above embodiment, after all the nodes Nn belonging to the previous group are won for content acquisition, the lottery process is performed by the nodes Nn belonging to the next group. However, the order of starting the content acquisition may be the order in which the line speed is fast, while the lottery processing time (content acquisition time) may be partially overlapped between the groups. In this case, in the period in which the content acquisition timing overlaps, the timing is matched between the groups. Therefore, it is necessary to match the waiting time to the group having the slowest line speed among the groups having the same period.

  In the above embodiment, each node Nn determines the content acquisition time by performing a lottery process based on the parameter information. However, another method may be used. For example, for each group, for example, the center server SA determines in advance which node Nn acquires the content in which order. Then, the center server SA creates parameter information in which the content acquisition order of the group (the faster the line speed is, the faster the group), the acquisition timing of each node in each group, the waiting time, etc. are set. Deliver to. Each node Nn has its own group content acquisition order, the waiting time of the group whose acquisition order is earlier than its own group, and how many times content acquisition is performed within each group, Can be requested. Each node Nn can obtain the content acquisition time of the own node from the order of the own node in the own group and the standby time of the own group.

  In addition, as the line speed used for grouping, the upstream bandwidth may be used, or the downstream bandwidth may be used. Further, the upstream bandwidth and the downstream bandwidth may be used together.

  Further, in the above embodiment, as shown in Expression 2, for example, in the lottery process, the winning number is determined based on the number of uploadable units, the number of nodes Nn that have already been won, and the like, and the winning probability is calculated. However, it is not limited to this. For example, each node Nn may acquire content at random with a predetermined winning probability for each group. However, content distribution is not completed unless all nodes Nn acquire the content. Therefore, there is a need for a method such that the winning probability is finally 1 or, when a certain point in time is reached, all the remaining nodes Nn not won in the group acquire contents.

  Further, in the above embodiment, when the node Nn acquires content, it first downloads a part of the content from a plurality of content holding nodes, and then downloads all of the content from the content holding node with the fastest receiving speed. However, the present invention is not limited to this (steps S76 to S78 shown in FIG. 11). For example, one content holding node may be selected from the beginning, and all content may be downloaded from the selected content holding node. However, the actual number of uploads of each content holding node approaches the number of uploadable units by downloading using the former method. Therefore, the uploading load is distributed according to the line speed. Therefore, it is more preferable to take the former method.

  In the above-described embodiment, each node Nn belonging to each group acquires content at any of the acquisition times that arrive at each standby time. However, the present invention is not limited to this. For example, all the nodes Nn to which it belongs may acquire content at the same time. In this case, for example, first, all the nodes Nn belonging to the group 1 having the fastest line speed acquire the content. After the standby time of group 1 has elapsed since the start of content acquisition by group 1, next, all nodes Nn belonging to group 2 with the second highest line speed acquire content. After the standby time of group 2 has elapsed since the start of content acquisition by group 2, next, all nodes Nn belonging to group 3 with the third highest line speed acquire the content. In this way, content is acquired one after another. For some groups, all nodes Nn acquire content at the same time, and for other groups, each node Nn acquires content at any of the acquisition times that arrive at each standby time. You may make it acquire.

3 IX
4a, 4b ISP
5a, 5b DSL line operator 6 FTTH line operator 7 Communication line 8 Network 9 Overlay network 11 Control unit 12 Storage unit 13 Communication unit 14 Bus 21 Control unit 22 Storage unit 23 Buffer memory 24 Decoder unit 25 Video processing unit 26 Display unit 27 Audio Processing unit 28 Speaker 29 Communication unit 30 Input unit 31 Bus Nn node SA Center server S Content distributed storage system

Claims (6)

In a content distribution system including a plurality of node devices that can communicate with each other via a network,
The node device is
As the acquisition order of the content of a plurality of groups in which the plurality of node devices are grouped according to the line speed of a line for connecting to the network of each of the plurality of node devices, acquisition is performed in order from the group with the highest line speed. Group information acquisition means for acquiring order information indicating that content is to be acquired and time information indicating a time interval for determining whether to acquire content for each group, wherein the time information indicates a shorter time interval as the line speed increases When,
Based on the order information and time information acquired by the group information acquisition means, the time indicated by the time information at the time corresponding to the group to which the node device comprising the group information acquisition means belongs among the plurality of groups. A determination means for probabilistically determining whether to acquire content for each time interval of the group belonging to the interval;
Content acquisition means for acquiring content when it is determined to be acquired by the determination means;
Content delivery system, characterized in that the Ru with a. A node device in a content distribution system having a plurality of node devices mutually communicable through a network,
As the acquisition order of the content of a plurality of groups in which the plurality of node devices are grouped according to the line speed of a line for connecting to the network of each of the plurality of node devices, acquisition is performed in order from the group with the highest line speed. Group information acquisition means for acquiring order information indicating that content is to be acquired and time information indicating a time interval for determining whether to acquire content for each group, wherein the time information indicates a shorter time interval as the line speed increases When,
Based on the order information and time information acquired by the group information acquisition means, the time indicated by the time information at the time corresponding to the group to which the node device comprising the group information acquisition means belongs among the plurality of groups. A determination means for probabilistically determining whether to acquire content for each time interval of the group belonging to the interval;
When it is determined to be acquired by the determination unit, a content acquisition unit that searches for a node device that can transmit the content by acquiring the content and acquires the content from the searched node device ;
Node device comprising: a. The node device according to claim 2,
The group information acquisition means further acquires transmission number information in which the number of transmissions that can be transmitted is set to a higher value for a group with a higher line speed, as transmission number information indicating the number of possible transmissions of content of each node device for each group. And
The determination means, based on the order information and the transmission number information acquired by the group information acquisition means, the number of node devices that can transmit the content in each group and the transmittable number , A node device , characterized in that content is determined to be acquired with a probability corresponding to the number of node devices that have not acquired the content among node devices belonging to a group . In the node device according to claim 2 or 3,
It is time determination means for determining whether it is time for the determination by the determination means based on the order information acquired by the group information acquisition means, and the earlier the acquisition order of the belonging group, It further comprises a timing determination means for accelerating the determination timing,
The node device characterized by probabilistically determining whether to acquire content when it is determined by the time determination unit that it is time to perform the determination by the determination unit. A computer of a node device in a content distribution system including a plurality of node devices that can communicate with each other via a network ,
As the acquisition order of the content of a plurality of groups in which the plurality of node devices are grouped according to the line speed of a line for connecting to the network of each of the plurality of node devices, acquisition is performed in order from the group with the highest line speed. Group information acquisition step for acquiring order information indicating that the content is acquired and time information indicating a time interval for determining whether to acquire content for each group, and indicating time interval indicating a shorter time interval as the line speed is higher When,
The time indicated by the time information at the time corresponding to the group to which the node device that executed the group information acquisition step belongs among the plurality of groups based on the order information and time information acquired by the group information acquisition step. A determination step of probabilistically determining whether to acquire content for each time interval of the group belonging to the interval;
When it is determined to be acquired by the determination step, a content acquisition step of searching for a node device capable of transmitting the content by acquiring the content and acquiring the content from the searched node device ;
The content acquisition processing program characterized by performing this . In a content distribution method in a content distribution system including a plurality of node devices that can communicate with each other via a network,
As the acquisition order of contents of a plurality of groups in which the plurality of node devices are grouped according to a line speed of a line for the node device to connect to the network of each of the plurality of node devices, the line speed is Acquisition of order information indicating that acquisition is performed in order from a fast group, and time information indicating a time interval for determining whether to acquire content for each group, and time information indicating a shorter time interval as the line speed increases. A group information acquisition process,
Based on the order information and time information acquired by the group information acquisition step, the node device, according to the group to which the node device that executed the group information acquisition step belongs among a plurality of groups, A determination step for probabilistically determining whether to acquire content for each time interval of the group belonging to the time interval indicated by the time information;
A content acquisition step of acquiring content when it is determined that the node device acquires by the determination step;
Content delivery method characterized in that it comprises a.

Images