JP2018147171A - Content distribution method, and content distribution system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To predict content requested by a terminal device.SOLUTION: A content distribution method which is executed by a content distribution system for distributing content includes: a step for acquiring information showing a social relationship which is established among clients constructing a social network among a plurality of clients using content distribution services; and a step for predicting content to be used among the clients constructing the social network on the basis of the information showing the social relationship acquired in the acquisition step.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a content distribution method and a content distribution system.

Services that distribute large volumes of content at high speed are provided by optimizing load balancing and traffic using a large number of servers and high-speed communication networks. This service is classified into at least two types.
The first is to connect each of a plurality of distributed cache servers via a high-speed communication network. Then, an optimal cache server is selected according to the traffic situation, and the requested content is distributed from the selected cache server to the terminal device used by the client.

Second, the terminal device requests content and downloads the requested content. Then, the terminal device that downloaded the content plays the role of a cache server, and in response to subsequent requests for the downloaded content, distributes the content in cooperation with other terminal devices that have already downloaded the content. To do. This service has a so-called P2P (Peer to Peer) configuration.
In the content distribution service, the terminal device acquires requested content from the origin server.
In the content distribution service, when there is a popular content, communication quality tends to deteriorate due to the traffic concentrated on the origin server that manages the content among a plurality of origin servers. For this reason, it is performed to determine the timing of content placement and the capacity of the content to be placed on the cache server.

  A user-selected image for selecting another user is displayed on the display unit based on the viewing information of the other user, one other user is selected from the user-selected image based on the user input, and the viewing information of the other user is selected. Based on this, a technique is known in which at least one of content that the selected other user has viewed in the past, content that is currently viewed or content that is scheduled to be viewed in the future is selected and displayed on a display unit. (For example, refer to Patent Document 1).

JP 2009-141952 A

Consider a case where a cache server delivers content requested by a terminal device. In this case, it is preferable that the content requested by the terminal device is predicted before the terminal device requests the content, and the predicted content is arranged in the cache server. However, a technology for predicting the content requested by the terminal device has not been established.
Even if a technology for predicting the content requested by the terminal device is established, in order to cache all the predicted content in each of the plurality of cache servers, a huge storage area is required in the cache server. Become.

  The present invention has been made in view of the above problems, and an object thereof is to predict content requested by a terminal device.

The present invention provides the following means in order to solve the above problems.
(1) A content distribution method according to a first aspect is a content distribution method executed by a content distribution system that distributes content, and a client that constructs a social network among a plurality of clients that use a content distribution service Content that is used between clients constructing the social network based on the step of acquiring information indicating the social relationship established between them and the information indicating the social relationship acquired in the step of acquiring Predicting the content.
(2) In the content distribution method according to the above aspect, the step of dividing the content predicted in the prediction step into a plurality of pieces, and the first piece of the plurality of pieces is arranged in each of the plurality of cache servers. And a step of distributing pieces other than the first piece to the plurality of cache servers.
(3) In the content distribution method according to the above aspect, a step of receiving a content distribution request transmitted by a terminal device of a request client that requests content, and a plurality of the plurality of the plurality of content distribution methods based on content identification information included in the content distribution request A step of connecting to one of the connection cache servers of the cache server, and a cache server other than the connection cache server connected in the connection step transmits the pieces other than the first piece to the connection cache server And the connection cache server distributes the first piece, the piece other than the first piece, and the piece other than the first piece transmitted by the other cache server to the terminal device. You may make it have.
(4) The content distribution method according to the above aspect includes a step of setting predetermined content, and in the step of acquiring, the society established in the social network constructed by the client who has requested the predetermined content You may make it acquire a target relationship.
(5) In the content distribution method according to the above aspect, detecting an influencer from a client that constructs the social network based on the information indicating the social relationship acquired in the acquiring step; Identifying the social network constructed by the influencer, and in the arranging step, the plurality of clients to which the client terminal device constructing the social network identified in the identifying step is connected The first piece may be arranged in each of the cache servers, and pieces other than the first piece may be distributed and arranged in each of the plurality of cache servers.
(6) In the content distribution method according to the above aspect, in the step of predicting the change of the social network specified in the specifying step, and the new social network obtained by predicting the change of the social network A step of detecting a new influencer and a step of identifying a social network constructed by the new influencer, wherein in the arranging step, a terminal device of a client constructing the other social network You may make it distribute | distribute and arrange | position to the said some cache server connected.
(7) The content distribution system according to the first aspect is a content distribution system that distributes content, and is a social network that is established among clients that construct a social network among a plurality of clients that use a content distribution service. An acquisition unit that acquires information indicating a relationship; a prediction unit that predicts content used between clients constructing the social network based on the information indicating the social relationship acquired by the acquisition unit; Is a content distribution system.

  According to the content delivery system concerning the said aspect, the content which a terminal device requests | requires can be estimated.

It is a figure showing an example of a contents distribution system concerning an embodiment. It is a functional block diagram which shows an example of the content delivery management apparatus which concerns on embodiment. It is a figure which shows an example of network management information. It is a figure which shows an example of content management information. It is a figure for demonstrating a tripartite closure. It is a flowchart which shows an example of operation | movement of the content delivery management apparatus which concerns on embodiment. It is a sequence chart which shows an example of operation | movement of the content delivery management apparatus which concerns on embodiment. It is a figure which shows an example of operation | movement of the content delivery system which concerns on embodiment. It is a figure which shows an example of the reduction amount of the cache data in the content delivery system which concerns on embodiment.

Hereinafter, the present embodiment will be described in detail with appropriate reference to the drawings.
(Embodiment)
FIG. 1 is a diagram illustrating an example of a content distribution system according to an embodiment.
The content distribution system 1 includes a terminal device 100-1, a terminal device 100-2, a terminal device 100-3, a cache server 200-1, a cache server 200-2, a cache server 200-3, an origin server 400, and a content distribution management device. 500.
The content distribution system 1 further includes a switch 300-1, a switch 300-2, a switch 300-3, a switch 300-4, a switch 300-5, a switch 300-6, a switch 300-7, and a switch 300-8. .
Hereinafter, an arbitrary terminal device among the terminal device 100-1, the terminal device 100-2, and the terminal device 100-3 is referred to as a terminal device 100. In addition, an arbitrary cache server among the cache server 200-1, the cache server 200-2, and the cache server 200-3 is referred to as a cache server 200. In addition, any switch among the switch 300-1, the switch 300-2, the switch 300-3, the switch 300-4, the switch 300-5, the switch 300-6, the switch 300-7, and the switch 300-8 is switched. 300.
The terminal device 100, the cache server 200, the switch 300, the origin server 400, and the content distribution management device 500 are connected via a communication network 50 such as the Internet. Furthermore, a content distribution network is configured by the plurality of switches 300.

The content distribution system 1 according to the present embodiment provides a content distribution service, which is a service for distributing content, to the terminal device 100 of the client.
The terminal device 100 is a device in which a client subscribed to a content distribution service requests content distribution.
The cache server 200 stores a copy of the content stored in the origin server 400 on the communication network 50, and distributes it in place of the origin server 400 when requested by the terminal device 100. Thereby, network traffic and server load can be distributed.
The switch 300 relays and transfers data between devices connected to the communication network 50.
The origin server 400 accumulates content.

  The content distribution management device 500 provides a content distribution service to the client terminal device 100. The content distribution management apparatus 500 includes Facebook (registered trademark), LINE (registered trademark), Twitter (registered trademark), mixi (registered trademark), Youtube (registered trademark), and Instagram, among a plurality of clients that use the content distribution service. Information indicating a social relationship established between a plurality of clients constituting a social networking service (SNS) such as (registered trademark) is acquired. Specifically, the content distribution management device 500 classifies a plurality of clients that use the content distribution service into communities. A plurality of clients belong to the community. Hereinafter, a community formed by classifying a plurality of clients using the content distribution service is referred to as a “first community”, and each of the plurality of clients belonging to the first community is referred to as a “first client”.

The content distribution management device 500 specifies the most popular content among a plurality of clients that use the content distribution service, and also provides information on the first client that uses the content (hereinafter referred to as “SNS information”). collect. Specifically, if a community is formed by Twitter, the content distribution management device 500 uses SNS information such as follower information of a specific first client of Twitter to display the popular content. Extract relationships that hold between the first clients used.
Furthermore, the content distribution management apparatus 500 identifies a person who has a high influence level among the first clients that use popular content. Hereinafter, the highly influenced person identified by the content distribution management apparatus 500 from the first client is referred to as a “first influencer”.
In addition, the content distribution management device 500 specifies a client belonging to a community other than the first community to which the first influencer belongs. Hereinafter, a community other than the first community to which the first influencer belongs is referred to as a “second community”. Then, the content distribution management device 500 identifies clients belonging to the second community. Hereinafter, a client belonging to the second community is referred to as a “second client”.

Furthermore, the content distribution management device 500 predicts future changes in the second community. The content distribution management device 500 identifies a person who has a high influence level based on the prediction result of future changes in the second community. Hereinafter, the highly influenced person identified by the content distribution management apparatus 500 from the second client under the prediction result of the future change of the second community is referred to as a “second influencer”.
In addition, the content distribution management device 500 specifies a client belonging to a community other than the second community to which the second influencer belongs. Hereinafter, a community other than the second community to which the second influencer belongs is referred to as a “third community”, and a client belonging to the third community is referred to as a “third client”.

The content distribution management device 500 specifies a plurality of cache servers to which the second client and the third client are connected. Hereinafter, each of the plurality of cache servers to which the second client and the third client are connected is referred to as a “connection cache server”.
The content distribution management device 500 acquires popular content from the origin server 400. Then, the content distribution management device 500 divides the acquired popular content, and disperses and arranges each of the plurality of contents obtained by the division to each of the plurality of connection cache servers. Hereinafter, each of a plurality of contents obtained by dividing popular content is referred to as “divided content”. Specifically, the content distribution management device 500 transmits a content request to the origin server 400 to request popular content. The content request includes the content ID of the content to be acquired.

The origin server 400 receives the content request transmitted from the content distribution management device 500, and specifies the content corresponding to the content ID included in the received content request. The origin server 400 transmits the designated content to the content distribution management device 500.
The content distribution management device 500 receives the content transmitted by the origin server 400 and creates the divided content by dividing the received content. Each of the plurality of divided contents includes a content ID of the divided content (hereinafter referred to as “divided content ID”) in addition to the divided content.
The content distribution management apparatus 500 arranges the divided content to be reproduced first among the divided contents on all of the plurality of connection cache servers. Hereinafter, the divided content to be reproduced first is referred to as “first reproduced divided content”. The content distribution management apparatus 500 creates the same number of first reproduction divided contents as the number of connection cache servers by duplicating, and creates each of the plurality of first reproduction division contents thus created for each of the plurality of connection cache servers. To place.
Furthermore, the content distribution management device 500 determines a connection cache server in which each of one or a plurality of divided contents other than the first reproduction divided contents is arranged based on the storage capacity of each connection cache server. The content distribution management server 500 transmits each of one or a plurality of divided contents with each determined connection cache server as a destination.

The first reproduction divided content and the divided content transmitted by the content distribution management device 500 are transmitted to the connection cache server 300 via one or a plurality of switches 300. The connection cache server 300 receives the first reproduction divided content or the first reproduction divided content and the divided content transmitted by the content distribution management device 500, and divides the received first reproduction divided content or the first reproduction division content. Store content.
The second client belonging to the second community to which the first influencer belongs and the third client belonging to the third community to which the second influencer belongs distribute the content by operating the terminal device 100. Request. By this operation, the terminal device 100 creates a content delivery request and transmits the created content delivery request to a connection cache server having a short distance among the connection cache servers. The content distribution request includes a content ID.

The connection cache server that has received the content distribution request distributes the first reproduction content of the content corresponding to the content ID to the terminal device 100 based on the content ID included in the received content distribution request. Hereinafter, the connection cache server that has received the content distribution request is referred to as an “edge server”.
While distributing the first reproduction divided content, the edge server transmits a divided content request including information requesting the divided content to be distributed next to the first reproduction divided content to the content distribution management device 500.
The content distribution management device 500 receives the divided content request and specifies a connection cache server that stores the divided content to be distributed next to the first reproduction divided content included in the received divided content request. The content distribution management device 500 transmits a divided content transfer request including information requesting to transfer the divided content distributed next to the first reproduction divided content to the edge server to the identified connection cache server. The divided content transfer request includes the IP address of the edge server and the divided content ID of the divided content to be distributed next to the first reproduction divided content.

The connection cache server that has received the divided content transfer request corresponds the divided content corresponding to the divided content ID to the IP address based on the IP address and the divided content ID of the edge server included in the received divided content transfer request. To the edge server.
The edge server receives the divided content transmitted by the connection cache server, and stores the received divided content. When the distribution of the first reproduction divided content is completed, the edge server distributes the stored divided content to the terminal device 100.
While distributing the divided content, the edge server transmits a divided content request including information requesting the divided content to be distributed next to the distributed content to be distributed to the content distribution management apparatus 500.
The content distribution management device 500 receives the divided content request and specifies a connection cache server that stores the divided content to be distributed next included in the received divided content request. The content distribution management device 500 transmits a divided content transfer request including information requesting to transfer the divided content to be reproduced next to the edge server to the identified connection cache server.

The connection cache server that has received the divided content transfer request corresponds the divided content corresponding to the divided content ID to the IP address based on the IP address and the divided content ID of the edge server included in the received divided content transfer request. To the edge server.
The edge server receives the divided content transmitted by the connection cache server, and stores the received divided content. When the distribution of the divided content ends, the edge server distributes the stored divided content to the terminal device 100.
The edge server repeats the same processing as described above until there is no divided content to be distributed next to the divided content.

(Content distribution management device)
FIG. 2 is a functional block diagram illustrating an example of a content distribution management apparatus according to the embodiment.
The content distribution management device 500 includes a communication unit 502, a storage unit 504, and a control unit 506.
The communication unit 502 communicates with the cache server 200, the switch 300, the origin server 400, and the like via the communication network 50. Specifically, the communication unit 502 transmits the content request output from the control unit 506 to the origin server 400. The communication unit 502 receives the content transmitted by the origin server 400 and outputs the received content to the content management unit 516.
Further, the communication unit 502 transmits the first reproduction divided content and the divided content output from the control unit 506 to the destination included in the first reproduction divided content and the divided content via the switch 300.

In addition, the communication unit 502 receives the divided content request transmitted by the edge server, and outputs the received divided content request to the control unit 506. The communication unit 502 transmits the divided content transfer request output from the control unit 506 to the connection cache server 200.
The storage unit 504 stores a program 5042, network management information 5044, and content management information 5046. The program 5042 causes the control unit 506 to function as the setting unit 508, the acquisition unit 510, the detection unit 512, the prediction unit 514, the content management unit 516, and the distribution unit 522. Further, the content management unit 516 functions as a dividing unit 518 and an arrangement unit 520.

(Network management information)
FIG. 3 is a diagram illustrating an example of network management information.
The network management information 5044 is information in a table format in which a cache server ID, an IP address, and a user ID of the terminal device 100 connected to the server capacity information are stored in association with each other for each of a plurality of cache servers. In the example shown in FIG. 3, the cache server ID “CS1001”, the IP address “192.168.10.10”, the server capacity “www”, and the user IDs “user001, user002, user003,. ing.
Further, the cache server ID “CS1002”, the IP address “192.168.10.11”, the server capacity “xxx”, and the user IDs “user001, user002, user003,...” Are associated with each other.
Further, the cache server ID “CS1003”, the IP address “192.168.10.12”, the server capacity “yyy”, and the user IDs “user001, user002, user003,...” Are associated with each other.

(Content management information)
FIG. 4 is a diagram illustrating an example of content management information.
The content management information 5046 includes, for each of a plurality of contents, a content ID, a divided content ID of each of the plurality of divided contents obtained by dividing the content, and a connection cache server ID of a connection cache server in which each divided content is arranged. Store it in association.
In the example shown in FIG. 4, the content ID “CON001” is associated with the divided content IDs “PIE001001”, “PIE001002”, and “PIE001003”. Further, “PIE001001”, “PIE001002”, and “PIE001003” are associated with the connection cache server IDs “All”, “CS1002”, and “CS1003”, respectively. Here, “ALL” indicates all connection cache servers.
Further, the content ID “CON002” and the divided content IDs “PIE002001”, “PIE002002”, “PIE002003”, and “PIE002004” are associated with each other. Furthermore, “PIE002001”, “PIE002002”, “PIE002003”, and “PIE002004” are associated with the connection cache server IDs “All”, “CS1001”, “CS1003”, and “CS1004”, respectively.
Further, the content ID “CON003” and the divided content IDs “PIE003001”, “PIE003002”, and “PIE002003” are associated with each other. Further, “PIE003001”, “PIE003002”, and “PIE003003” are associated with the connection cache server IDs “All”, “CS1003”, and “CS1001”, respectively.

Returning to FIG.
The control unit 506 is realized by an arithmetic device such as a CPU (Central Processing Unit). The control unit 506 functions as the setting unit 508, the acquisition unit 510, the detection unit 512, the prediction unit 514, the content management unit 516, and the distribution unit 522 by executing the program 5042 stored in the storage unit 504. Further, the control unit 506 functions as a dividing unit 518 and an arrangement unit 520.
The setting unit 508 specifies the most popular content in a predetermined cycle from all the content used in the content distribution system, and sets the specified content. The setting unit 508 outputs the content ID of the set content to the content management unit 516. Specifically, the setting unit 508 acquires the number of circulations on the SNS, and specifies the most popular content based on the acquired number of circulations on the SNS.

  The acquisition unit 510 classifies a plurality of clients that use the content distribution service into the first community. And the acquisition part 510 acquires the SNS information regarding the 1st client using the most popular content which the setting part 508 set among the 1st clients contained in a 1st community. Specifically, the acquisition unit 510 accesses the most popular content by searching for the SNS based on information such as the URL of the most popular content and the title of the most popular content. Identify the client who went. The acquisition unit 510 collects follower information and follower information of the identified client.

The detection unit 512 identifies the first influencer from the first client included in the first community classified by the acquisition unit 510. The detection unit 512 identifies the second community to which the identified first influencer belongs, and identifies the second client belonging to the identified second community. The detection unit 512 outputs the user ID of the second client to the content management unit 516. Specifically, the detection unit 512 creates a social graph based on each of the plurality of first clients and the follower information of the first client, and extracts a second community. The detection unit 512 outputs information indicating the extracted second community to the prediction unit 514.
Alternatively, the detection unit 512 determines a target user using a ranking method such as a HITS algorithm. The detection unit 512 generates a network between target users by forming a link when mutual follow is recognized between the users. The detection unit 512 extracts a cluster using a method of analyzing a network such as a CNM (Clauset. Newman Moore) algorithm, and performs clan extraction with n-Clan (n = 2) for each extracted cluster. And the clan extracted there may be used as a community.
Furthermore, the detection unit 512 specifies the second influencer based on the prediction result of the future change of the second community output by the prediction unit 514 described later. The detection unit 512 identifies the third community to which the second influencer belongs, and identifies the third client belonging to the identified third community. The detection unit 512 outputs the user ID of the third client to the content management unit 516. By identifying the second influencer, it is possible to identify a node with a high power center that is related to the first influencer and can soon become an influencer.

The prediction unit 514 predicts the future of the second community based on the information indicating the second community output from the detection unit 512. The prediction unit 514 predicts future changes in the second community based on the strength of the relationship between the second clients in the second community detected by the detection unit 512. The prediction unit 514 outputs the prediction result of the future change of the second community to the detection unit 512. Specifically, the prediction unit 514 uses one or both of three-way closure and homophilicity to form a new graph by predicting a link that is likely to generate a new relationship, Predicting future changes in the second community by performing community extraction on a simple graph. Here, the tripartite closure and the same kind of affinity will be described.
(Triple closure)
FIG. 5 is a diagram for explaining an example of a three-party closure. In FIG. 5, A, B, and C indicate users, and the line segments connecting the users are related to each other.
Three-party closure means that if user A and user B have a friendship and user A and user C have a friendship in a social network, there is a high possibility that user B and user C will have a friendship in the future. It is the feature. That is, it shows that those who have a common friend are likely to become friends in the future. Using this feature, a new link (link connecting user B and user C) is added to the social network.
(Homogeneous affinity)
Homogeneous affinity means the property that a relationship is likely to be created between similar nodes in a social network. Originally, the similarity of nodes needs to be determined in consideration of node attributes such as profiles and past activities. However, in the content distribution system according to the embodiment, two nodes are interchanged when viewed from the entire graph. Also, nodes having similar centrality are determined as similar nodes. Add a new link between similar nodes.

Returning to FIG.
The dividing unit 518 of the content management unit 516 creates a content request including the content ID output by the setting unit 508 and outputs the content request to the communication unit 502. In response to the content request, the content management unit 516 acquires the content output from the communication unit 502 and divides the acquired content to create a plurality of divided content. Specifically, the dividing unit 518 divides the content into a number equal to or less than the sum of the cache server connected to the second client and the cache server connected to the third client. The dividing unit 518 outputs a plurality of divided contents to the arranging unit 520.

The placement unit 520 acquires the user ID of the second client and the user ID of the third client output from the detection unit 512, and sets the acquired user ID of the second client and the user ID of the third client. Based on the network management information 5044, the cache server ID associated with the user ID of the second client and the user ID of the third client is acquired. Furthermore, the arrangement unit 520 acquires an IP address and server capacity information corresponding to each acquired cache server ID.
Furthermore, the arrangement unit 520 determines a connection cache server that arranges each of the plurality of divided contents output by the division unit 518 based on the server capacity. Specifically, the first reproduction divided content is arranged on all of the plurality of cache servers connected to the second client and the plurality of cache servers connected to the third client. Then, the arrangement unit 520 arranges divided contents other than the first reproduction divided contents in order from the connection cache server having the largest server capacity.
The arrangement unit 520 stores the content ID, the plurality of divided content IDs, and the cache server ID of the connected cache server in which each of the plurality of divided contents is arranged in association with the content management information 5046. Arrangement unit 520 outputs each of a plurality of divided contents addressed to the IP address of the connection cache server to be arranged to communication unit 502.

  The distribution unit 522 acquires the divided content request output from the communication unit 502 and, based on the divided content ID included in the acquired divided content request, from the content management information 5046 stored in the storage unit 5046, the divided content. The cache server ID associated with the ID is acquired. Then, the distribution unit 522 acquires an IP address corresponding to the acquired cache server ID from the network management information 5044 stored in the storage unit 5046. The distribution unit 522 creates a divided content transfer request including the divided content ID with the acquired IP address as the destination, and outputs the divided content transfer request to the communication unit 502.

(Operation of content distribution management device)
FIG. 6 is a flowchart illustrating an example of the operation of the content distribution management apparatus.
(Step S102) The setting unit 508 of the content distribution management device 500 identifies the most popular content from all the content used in the content distribution system, and sets the identified content.
(Step S104) The acquisition unit 510 of the content distribution management device 500 classifies a plurality of clients that use the content distribution service into the first community. The acquisition unit 510 acquires SNS information related to the first client using the content set by the setting unit 508 among the first clients included in the first community.
(Step S106) The detection unit 512 of the content distribution management device 500 specifies the first influencer from the first clients included in the first community classified by the acquisition unit 510.

(Step S108) The detection unit 512 of the content distribution management device 500 identifies the second community to which the first influencer belongs, and identifies the second client belonging to the identified second community.
(Step S110) The prediction unit 514 of the content distribution management device 500 predicts the future of the second community based on the information indicating the second community output from the detection unit 512. The detection unit 512 identifies the second influencer under the prediction result of the future change of the second community.
(Step S112) The detection unit 512 of the content distribution management device 500 identifies the third community to which the second influencer belongs, and identifies the third client belonging to the identified third community.

(Step S <b> 114) The dividing unit 518 of the content distribution management apparatus 500 creates a content request including the content ID output by the setting unit 508 and outputs the content request to the communication unit 502. In response to the content request, the dividing unit 518 acquires the content output from the communication unit 502 and divides the acquired content to create a plurality of divided contents.
The placement unit 520 acquires the information indicating the second client and the information indicating the third client output from the detection unit 512, and acquires the acquired information indicating the second client and the information indicating the third client. Based on the network management information 5044, a plurality of cache server IDs connected to the second client and a plurality of cache server IDs connected to the third client are acquired. Furthermore, the arrangement unit 520 acquires the IP address and server capacity information corresponding to each of the plurality of cache server IDs from the content management information 5046. The placement unit 520 determines a connection cache server that places each of the plurality of divided contents output from the division unit 518 based on the server capacity.
The arrangement unit 520 stores the content ID, the plurality of divided content IDs, and the cache server ID of the connected cache server in which each of the plurality of divided contents is arranged in association with the content management information 5046.

(Operation of content distribution system)
The operation of the content distribution system according to the embodiment will be described with reference to FIGS. 7 and 8.
FIG. 7 is a sequence chart showing an example of the operation of the content distribution system according to the embodiment.
In the example shown in FIG. 7, the content distribution management device 500 divides the content into four parts, and each of the four divided contents is assigned to the cache server 200-1, the cache server 200-2, and the cache server 200-3. The operation | movement after arrange | positioning to the cache server 200-4 is shown.

FIG. 8 is a diagram illustrating an example of a content distribution system. The content distribution system 1 includes a plurality of connection cache servers 200 distributed on the communication network 50 and a content distribution management device 500.
The content distribution management device 500 arranges the first reproduction divided content (Piece (0)) in the cache server 200-1. In addition, the content distribution management device 500 arranges the first reproduction divided content (Piece (0)) and the divided content (Piece (1)) in the cache server 200-2. In addition, the content distribution management device 500 arranges the first reproduction divided content (Piece (0)) and the divided content (Piece (2)) in the cache server 200-3. In addition, the content distribution management device 500 arranges the first reproduction divided content (Piece (0)) and the divided content (Piece (3)) in the cache server 200-4.
Returning to FIG. 7, the description will be continued.
(Step S202) The terminal device 100 transmits a content distribution request to the cache server 200-1 that is close in distance.
(Step S <b> 204) The cache server 200-1 distributes the first reproduction divided content of the content corresponding to the content ID included in the content distribution request transmitted by the terminal device 100 to the terminal device 100.
(Step S206) The cache server 200-1 distributes the first reproduction divided content (Piece (0)) to the content distribution management device 500 while distributing the first reproduction divided content (Piece (0)). A split content request including information requesting the split content (Piece (1)) is transmitted.

(Step S <b> 208) Upon receiving the divided content request transmitted from the cache server 200-1, the communication unit 502 of the content distribution management device 500 outputs the received divided content request to the control unit 506. The distribution unit 522 of the control unit 506 uses the cache server 200-2 storing the divided content (Piece (1)) based on information requesting the divided content (Piece (1)) included in the divided content request. A divided content transfer request as a destination is created, and the created divided content transfer request is output to the communication unit 502. The communication unit 502 transmits the divided content transfer request output from the distribution unit 522 to the cache server 200-2.
(Step S210) Upon receiving the divided content transfer request transmitted by the content distribution management device 500, the cache server 200-2 sends the divided content (Piece (1) to the cache server 200-1 in response to the received divided content transfer request. )) Forward. The cache server 200-1 stores the divided content (Piece (1)) transmitted by the cache server 200-2.
(Step S212) When the distribution of the first reproduction divided content is completed, the cache server 200-1 continues to distribute the divided content (Piece (1)) to the terminal device 100.
(Step S214) While distributing the divided content (Piece (1)), the cache server 200-1 sends to the content distribution management device 500 the next divided content (Piece (Piece (1))). 2)) A split content request including information requesting is transmitted.

(Step S216) Upon receiving the divided content request transmitted from the cache server 200-1, the communication unit 502 of the content distribution management device 500 outputs the received divided content request to the control unit 506. The distribution unit 522 of the control unit 506 selects the cache server 200-3 storing the divided content (Piece (2)) based on the information requesting the divided content (Piece (2)) included in the divided content request. A divided content transfer request as a destination is created, and the created divided content transfer request is output to the communication unit 502. The communication unit 502 transmits the divided content transfer request output from the distribution unit 522 to the cache server 200-3.
(Step S218) In response to the divided content transfer request transmitted by the content distribution management device 500, the cache server 200-3 transfers the divided content (Piece (2)) to the cache server 200-1. The cache server 200-1 stores the divided content (Piece (2)) transmitted by the cache server 200-3.
(Step S220) When distribution of the divided content (Piece (1)) is completed, the cache server 200-1 continues to distribute the divided content (Piece (2)) to the terminal device 100.

(Step S222) While distributing the divided content (Piece (2)), the cache server 200-1 sends to the content distribution management device 500 the next divided content (Piece (2)) of the divided content (Piece (2)). 3)) A split content request including information requesting is transmitted.
(Step S224) Upon receiving the divided content request transmitted from the cache server 200-1, the communication unit 502 of the content distribution management apparatus 500 outputs the received divided content request to the control unit 506. The distribution unit 522 of the control unit 506 selects the cache server 200-3 storing the divided content (Piece (3)) based on the information requesting the divided content (Piece (3)) included in the divided content request. A divided content transfer request as a destination is created, and the created divided content transfer request is output to the communication unit 502. The communication unit 502 transmits the divided content transfer request output from the distribution unit 522 to the cache server 200-4.
(Step S226) The cache server 200-4 transfers the divided content (Piece (3)) to the cache server 200-1 in response to the divided content transfer request transmitted by the content distribution management device 500. The cache server 200-1 stores the divided content (Piece (3)) transmitted by the cache server 200-4.
(Step S228) When distribution of the divided content (Piece (2)) is completed, the cache server 200-1 continues to distribute the divided content (Piece (3)) to the terminal device 100.
At this time, the edge server stores the first reproduction divided content, the divided content (Piece (1)), the divided content (Piece (2)), and the divided content (Piece (3)). After a certain period of time, the divided content (Piece (1)), the divided content (Piece (2)), and the divided content (Piece (3)) are deleted. With this configuration, the amount of data cached in the edge server can be reduced.

According to the sequence chart shown in FIG. 7, the content distribution system provides the first reproduction divided content or the first reproduction divided content and the divided content to each of the plurality of connection cache servers 200 that distribute the content to the terminal device 100. To cache. Then, among the plurality of connection cache servers 200, the edge server acquires the divided content to be distributed next while distributing the first reproduced divided content or divided content, and obtains the first reproduced divided content or divided content. After the distribution is completed, the acquired divided content to be reproduced next is distributed. By configuring in this way, it is possible to speed up the content distribution as in the case where the content is fully cached in one cache server.
Here, until the edge server completes the distribution of the divided content (piece (k)) (k is an integer) to the terminal device 100, the next divided content (piece (k + 1)) is acquired in time. Consider the case where it was assumed that there was not. In this case, a time during which content cannot be distributed to the terminal device 100 occurs, and content distribution delay occurs. That is, the edge server cannot distribute the content until the distribution of the divided content (piece (k + 1)) is started after the distribution of the divided content (piece (k)) is completed.

In order to prevent delays in content distribution, based on the communication speed between the edge server and the connection cache server other than the edge server and the communication speed between the terminal device 100 and the edge server, You need to set the size.
Hereinafter, an example of setting the size of the divided content will be described.
If the size of the divided content (piece (k)) is Sk [bit], Expression (1) is established.

  Sk = T ^ k * S * ((1-T) / (1-T ^ N)) (k = 0, 1,..., N-1) (1)

In Expression (1), “N” is the number of divided contents, and “S” is the size [bit] of the contents. “T” is b divided by a (b / a), and “a” is the communication speed between each cache server including an edge server and a terminal device that requests content from each cache server. The highest communication speed [bps], and “b” is the lowest communication speed [bps] of communication speeds between any two cache servers.
In other words, by dividing the content with the divided content size “Sk” obtained by the expression (1), it is possible to prevent a delay in content distribution and to reduce the size of the first reproduction divided content.

  The edge server and the cache server other than the edge server cache the first reproduction divided content of the content requested by the terminal device 100. As a result, it is possible to reduce the amount of data cached in one cache server while achieving high-speed content distribution equivalent to the case where all data of content is cached in one cache server. In this case, the cache data reduction amount for one content is obtained by the following equation (2) obtained by subtracting the size of the first reproduction divided content from the size of the entire content.

  S−S × ((1-T) / (1-T ^ N)) (2)

FIG. 9 is a diagram illustrating a reduction amount of cache data for one content. In FIG. 9, the X axis (horizontal axis) indicates the value of T (0 <T <1), and the Y axis (vertical axis) indicates the data reduction amount (0 <D <100) [%].
As shown in FIG. 9, the reduction amount of the cache data for one content shown in Expression (2) varies depending on the value of T. In the content distribution system according to the embodiment, the reduction amount of cache data for one content monotonously increases as T increases. In other words, as the difference between the communication speed at which the edge server acquires the divided content and the communication speed at which the edge server transmits the divided content becomes smaller, the size of the first reproduction divided content becomes smaller, and the cache data Reduction amount becomes large. It can also be seen that the number of divided contents “N” increases monotonically as T increases.

In the embodiment described above, the content distribution system includes the terminal device 100-1, the terminal device 100-2, the terminal device 100-3, the cache server 200-1, the cache server 200-2, the cache server 200-3, and the origin server 400. However, the present invention is not limited to this example. For example, the number of terminal devices 100 may be one or two, or four or more. The cache server 200 may be one or two, or four or more. Two or more origin servers 400 may be provided.
In the embodiment described above, the content distribution system includes the switch 300-1, the switch 300-2, the switch 300-3, the switch 300-4, the switch 300-5, the switch 300-6, the switch 300-7, and the switch 300-8. However, the present invention is not limited to this example. For example, the switch 300 may be one to seven, or nine or more.

In the above-described embodiment, when a community is configured with Twitter, the relationship established between the first clients using popular content using the follower information of the specific first client of the Twitter. However, the present invention is not limited to this example. For example, if a community is composed of social networking services such as Facebook, LINE, Twitter, mixi, YouTube, Instagram, etc., using the SNS information of the specific first client of that social networking service, it is popular A relationship that holds between the first clients that use the content may be extracted.
In the above-described embodiment, the case has been described in which the most popular content is specified by the number of circulations on the SNS, but is not limited to this example. For example, the setting unit 508 may identify the most popular content from content usage history data possessed by a content provider contracted with the content distribution service.

In the above-described embodiment, a case has been described in which a plurality of connection cache servers to which each of a plurality of second clients is connected match, but the present invention is not limited to this example. For example, the plurality of connection cache servers to which each of the plurality of second clients is connected may be different. In this case, the arrangement unit 520 extracts a common connection cache server among a plurality of connection cache servers to which each of the plurality of second clients is connected. The arrangement unit 520 may determine a connection cache server in which each of the plurality of divided contents is arranged based on the extracted server capacity of the connection cache server.
In the above-described embodiment, the case has been described where divided content other than the first reproduction divided content is arranged in order from the connection cache server having the largest server capacity, but is not limited to this example. For example, the arrangement unit 520 may determine a connection cache server that arranges the divided content at random between the connection cache servers having a server capacity capable of storing the divided content.

In the above-described embodiment, the case where the content distribution management device 500 specifies the most popular content has been described. However, the present invention is not limited to this example. For example, the content distribution management device 500 finds a community by performing a social network (SN) analysis on the SNS information. Then, the content distribution management apparatus 500 identifies the influencer of the found community, and predicts the content that members of other communities will use in the future from the history of the content used by the identified influencer. Good.
In the above-described embodiment, the case where the content distribution management apparatus 500 determines the connection cache server in which each of the plurality of divided contents is arranged based on the server capacity has been described, but the present invention is not limited to this example. For example, the content distribution management apparatus 500 may determine a connection cache server in which each of the plurality of divided contents is arranged based on the network speed, the cache storage area availability, the server load, and the like.

According to the content distribution system according to the embodiment, the content distribution management device 500 includes an influencer from one or a plurality of first clients using a certain content among a plurality of first clients included in the first community. And a plurality of second clients belonging to the second community to which the identified influencer belongs. Furthermore, the content distribution management device 500 specifies a plurality of third clients to which the third community to which the second influencer specified under the prediction of future changes in the second community belongs. Then, the content distribution management device 500 is a popular content from one or more cache servers connected to one or more second clients and one or more cache servers connected to a plurality of third clients. Select multiple cache servers to cache.
In other words, it is expected to be formed based on the cache server to which the second client belonging to the second community related to the first client using the popular content is connected and the future change of the second community. A cache server that caches a certain content is selected from cache servers to which a third client belonging to the third community connects.
The second client belongs to the second community to which the influencer identified by the first client belongs, and the third client is a third community that is expected to be formed based on future changes in the second community Therefore, the second client and the third client are assumed to have similar hobbies and preferences to the first influencer. For this reason, it is assumed that the second client and the third client have a high probability of using a certain content. That is, it can be said that the second client and the third client are predicted to use the certain content.
A plurality of cache servers included in the content distribution network by selecting a cache server for caching a certain content from a cache server to which a second client connects and a cache server to which a third client belonging to a third community connects The cache hit rate for each of the above can be improved. As a result, the average content distribution time of the content distribution network can be shortened.
According to the content distribution system according to the embodiment, a plurality of clients using the content distribution service are classified into communities based on the SNS analysis result. Then, a part of the content used by at least one member belonging to the community is arranged in advance via a content distribution network on a cache server connected to a terminal device of a user belonging to the same community. With this configuration, when a content request is actually generated, the content download time can be shortened, and the cache capacity used by each cache server can be reduced as much as possible. In addition, the content to be used can be predicted without specially configuring a device for predicting the content.

  Although the embodiments of the present invention have been described, these embodiments are presented as examples, and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and modifications thereof are included in the invention described in the claims and equivalents thereof, as long as they are included in the scope and gist of the invention.

  The terminal device, cache server, and content distribution management device described above may be realized by a computer. In that case, a program for realizing the function of each functional block is recorded on a computer-readable recording medium. The program recorded on the recording medium may be read by a computer system and executed by a CPU (control unit). The “computer system” here includes hardware such as an OS (Operating System) and peripheral devices. The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM. The “computer-readable recording medium” includes a storage device such as a hard disk built in the computer system.

  Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program. Further, the computer-readable recording medium may include a recording medium that dynamically holds the program for a short time. What holds the program dynamically for a short time is, for example, a communication line when the program is transmitted via a network such as the Internet or a communication line such as a telephone line.

In addition, the “computer-readable recording medium” may include a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that serves as a server or a client. The program may be for realizing a part of the functions described above. Further, the program may be a program that can realize the above-described functions in combination with a program already recorded in the computer system. The program may be realized using a programmable logic device. The programmable logic device is, for example, an FPGA (Field Programmable Gate Array).
The program may be for realizing a part of the functions described above.
Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.
In the above-described embodiment, the SNS is an example of a social network, the first reproduction divided content is an example of the first piece, the divided content is an example of the piece, and the content ID is an example of content identification information. An edge server is an example of a connection cache server.

100-1, 100-2, 100-3 ... Terminal device, 200-1, 200-2, 200-3, 200-4 ... Cache server, 300-1, 300-2, 300-3, 300-4, 300-5, 300-6, 300-7, 300-8 ... switch, 400 ... origin server, 500 ... content distribution management device, 502 ... communication unit, 504. Storage unit 506 ... Control unit 508 ... Setting unit 510 ... Acquisition unit 512 ... Detection unit 514 ... Prediction unit 516 ... Content management unit 518 Division unit, 520... Arrangement unit, 522. Distribution unit, 5042... Program, 5044... Network management information, 5046.

Claims (7)

A content distribution method executed by a content distribution system for distributing content,
Obtaining information indicating a social relationship established between clients that construct a social network among a plurality of clients using the content distribution service;
Predicting content to be used between clients constructing the social network based on the information indicating the social relationship acquired in the acquiring step. Dividing the predicted content in the predicting step into a plurality of pieces;
The first piece of the plurality of pieces is arranged in each of a plurality of cache servers, and pieces other than the first piece are distributed to the plurality of cache servers. The content delivery method described. Receiving a content distribution request transmitted by a terminal device of a requesting client that requests content;
Connecting to one of the plurality of cache servers based on content identification information included in the content distribution request;
A cache server other than the connected cache server connected in the connecting step transmits the pieces other than the first piece to the connected cache server;
The connection cache server distributes the first piece, the piece other than the first piece, and the piece other than the first piece transmitted by the other cache server to the terminal device; The content distribution method according to claim 2. Setting predetermined content, and
The content according to any one of claims 1 to 3, wherein, in the obtaining step, a social relationship established in the social network constructed by a client that has requested the predetermined content is obtained. Delivery method. Detecting an influencer from a client constructing the social network based on the information indicating the social relationship acquired in the acquiring step;
Identifying a social network built by the influencer,
In the arranging step, the first piece is arranged in each of the plurality of cache servers to which the client terminal devices that construct the social network identified in the identifying step are connected, and the plurality of cache servers The content delivery method according to claim 2, wherein pieces other than the first piece are distributed to each of the first and second pieces. Predicting changes in the social network identified in the identifying step;
Detecting a new influencer in a new social network obtained by predicting changes in the social network;
Identifying a social network built by the new influencer,
The content distribution method according to claim 5, wherein in the arranging step, the plurality of cache servers to which the terminal devices of the clients that construct the other social network are connected are arranged. A content distribution system for distributing content,
An acquisition unit that acquires information indicating a social relationship established between clients that construct a social network among a plurality of clients that use the content distribution service;
A content distribution system comprising: a prediction unit that predicts content used between clients constructing the social network based on information indicating the social relationship acquired by the acquisition unit.

Images