KR20130033253A - Overlay multicast system and its method to provide multiple content distribution in distributed content nodes - Google Patents

Abstract

PURPOSE: An overlay multicast system for distributing multiple contents in a regional distributed content node and a method thereof are provided to reduce network traffics and to improve the transmission speed and stability when distributing multiple contents through the content nodes distributed and arranged in the region. CONSTITUTION: An overlay management block provides resource context data to a context data management block by monitoring the dynamic data of the nodes(401). A service requester requests a service routing block 1 to create a multicast group(402). The service routing block 1 generates a group ID and selects a root node for a corresponding group(403). The service routing block 1 requests a content transmission block 1 to perform multicasting(412). The content transmission block 1 requests an identity management block to renew a multicast DB(413). The content transmission block 1 transmits multicast contents to a content transmission block 2(414). The service requestor requests the service routing block 1 to release the multicast group(415). The service routing block 1 requests the identity management block to release the multicast group(416). The identity management block deletes a corresponding group DB(417). [Reference numerals] (400) Requesting and receiving network context information; (401) Requesting and receiving resource context information; (402,404) Requesting to create a multicast group; (403) Generating a group ID and selecting a root node; (405) Generating a group DB; (406) Requesting to join or leave the multicast group; (407) Delivering the request for joining or leaving the multicast group; (408) Updating the group DB; (409) Requesting and receiving context information; (410) Requesting multicast channel setting; (411) Delivering the request for setting a multicast channel; (412) Requesting a multicast; (413) Requesting the update of a multicast DB; (414) Transmitting a multicast content; (415,416) Requesting the release of the multicast group; (417) Deleting the group DB; (AA) Service requester; (BB) Service routing block 1; (CC) Identity management block; (DD) Content delivery block 1; (EE) Service policy determination block; (FF) Context information management list; (GG) Overlay management block; (HH) Service routing block 2; (II) Content delivery block 2;

Description

Overlay multicast system and its method to provide multiple content distribution in distributed content nodes

The present invention relates to an overlay multicast system and method thereof, and more particularly to content pre-position, distributed cache update, software update (SW update) via distributed and locally located content nodes. The present invention relates to an overlay multicast system and method for reducing network traffic and improving transmission speed and stability when transmitting data to multiple nodes at the same time, such as instant message transmission and group streaming.

In the prior art, in order to simultaneously transmit content from one source node to several service nodes, IP (Internet Protocol) multicast technology is generally used. For example, IP multicast technology is applied to IPTV (Internet Protocol Television) service, which is a representative multicast service. Although IP multicast technology is known as an effective multicast technology, widespread practical use is delayed due to technical and policy limitations.

Therefore, a situation in which a universal service is not provided by using IP multicast technology on a commercial Internet, and an application layer multicast technology has been studied as an alternative.

In the application layer multicast scheme, each participant's end systems form a virtual network using a unicast connection, and end systems located on the virtual network configure a data transmission tree. When an end system receives a packet in the data transmission tree, the end system not only forwards the packet to its own application program, but also forwards the packet to end systems corresponding to child nodes of the transmission tree. In this way, the effect of multicasting can be seen using unicast at the network end.

Application layer multicast, which implements multicast technology as an application layer, mainly uses a multicast tree. Application layer multicast, similar to IP multicast, uses source nodes and entire receivers (receiving terminals) for data transmission. It is common to form a spanning tree between which the source node becomes the root node. However, this tree structure has some disadvantages as follows.

First, the throughput felt by lower nodes may be degraded by the network environment or nodes having poor computing performance among the middle nodes of the tree.

Second, since each receiver (receiving terminal) needs to receive data from only one parent node, the failure of the parent node has a direct and chain effect on the child nodes such as data transmission failure and delay.

As described above, the transmission performance of application layer multicast depends mainly on the topology of the multicast tree and the load of the sending server. Since the application layer has no knowledge of the network layer information, traffic congestion, node failure, There is a problem that can not effectively cope with network problems such as transmission delay, link failure, and can not guarantee the transmission quality.

In addition, the related art is a structure in which each node receives a file hierarchically from a top node in a tree structure and transmits the file to its own lower node. There is a problem that the transmission delay increases, and the E2E (end-to-end) transmission delay of the entire overlay network for sequential transmission for each layer increases.

Therefore, it is a problem of the present invention to solve the problems of the prior art as described above.

Accordingly, an object of the present invention is to provide an overlay multicast system and method for reducing network traffic and improving transmission speed and stability when distributing multiple contents through distributed content nodes arranged in a local area.

The objects of the present invention are not limited to the above-mentioned objects, and other objects and advantages of the present invention which are not mentioned can be understood by the following description, and will be more clearly understood by the embodiments of the present invention. It will also be readily apparent that the objects and advantages of the invention may be realized and attained by means of the instrumentalities and combinations particularly pointed out in the appended claims.

In order to achieve the above object, the present invention provides an overlay multicast system, comprising: a service policy decision block (SPD FE) for monitoring a lower network and a node to convey context information to a context information management block; The context information management block (CIM FE) for transferring the context information received from the service policy decision block to a service routing block; The service routing block (SR FE) for delivering the context information received from the context information management block to a content delivery block and for delivering a content delivery request to the content delivery block; And a content delivery block (CD) for establishing a multicast channel using the context information received from the service routing block, and delivering the corresponding content through the set multicast channel according to a content delivery request from the service routing block. FE).

Here, the content delivery block further performs a function of resetting a multicast channel by obtaining context information from the service routing block when a node failure occurs or a multicast group joins or leaves a party.

The content delivery block includes a group DB (Database) for managing multicast group members; And a multicast DB for managing the multicast request information.

The content delivery block manages a sequence number indicating a sequence number of the chunk file obtained by dividing the corresponding content into chunks.

On the other hand, the method of the present invention for achieving the above object, the overlay multicast method, the service policy decision block to monitor the lower network and the node to transfer the context information to the context information management block; Transmitting, by the context information management block, context information received from the service policy determination block to a service routing block; Delivering, by the service routing block, context information received from the context information management block to a content delivery block; Establishing, by the content delivery block, a multicast channel using the context information received from the service routing block; The service routing block forwarding a content delivery request to the content delivery block; And delivering, by the content delivery block, the corresponding content through the established multicast channel according to a content delivery request from the service routing block.

In addition, the method of the present invention further includes the step of reestablishing a multicast channel by the content delivery block obtaining context information from the service routing block in the event of a node failure or multicast group join or leave.

On the other hand, another system of the present invention for achieving the above object is an overlay multicast system, the context information management block (CIM FE) for requesting and obtaining the context information from the service policy decision block; Service routing for acquiring context information from the context information management block and delivering it to a content delivery block, selecting a group ID and a root node according to a multicast group creation request, and requesting the content delivery block to generate a multicast group. Block SR FE; Create a group DB according to a multicast group creation request from the service routing block, update the group DB according to a multicast group join or leave request, and set up a multicast channel using context information from the service routing block. And updating the multicast DB using the multicast request information according to the multicast content transmission request from the service routing block, and transmitting the content to a lower content delivery block through the set multicast channel. (CD FE).

Here, the content delivery block further performs a function of resetting a multicast channel by obtaining context information from the service routing block when a node failure occurs or a multicast group joins or leaves a party.

The content delivery block further performs a function of deleting the group DB according to a multicast group release request from the service routing block.

The content delivery block divides the content into chunks and divides the content into content delivery blocks in the same cluster, and updates content metadata indicating a location per chunk file (each node to which each chunk file is delivered). The content delivery blocks in the same cluster identify the location of the chunk file using the content metadata and exchange chunk files with each other using multiple nodes.

On the other hand, another method of the present invention for achieving the above object, the overlay multicast method, the context information management block to obtain the context information from the service policy decision block; The service routing block selecting a group ID and a root node according to the multicast group creation request and requesting the content delivery block to generate the multicast group; Generating, by the content delivery block, a group DB and updating the group DB according to a request for joining or leaving a multicast group from the service routing block; Obtaining, by the service routing block, context information from the context information management block and delivering the context information to the content delivery block; Establishing, by the content delivery block, a multicast channel using context information; And updating the multicast DB using the multicast request information according to the multicast content transmission request from the service routing block, and transmitting the content to the lower content delivery block through the set multicast channel. Steps.

In addition, the other method of the present invention further includes the step of resetting the multicast channel by obtaining the context information from the service routing block by the content delivery block when a node failure occurs or a multicast group joining or leaving a group occurs. .

Further, the other method of the present invention further includes the content delivery block deleting the group DB according to a multicast group release request from the service routing block.

As such, the present invention relates to an application layer multicast technology for providing efficient and fast transmission quality when transmitting the same content to a plurality of members on a geographically distributed content node, and each node that is a problem in the existing application layer multicast technology. It is possible to increase the transmission speed and stability by reducing the influence of subordinate node transmission due to the load, failure, and network transmission quality. In addition, to reduce the overall end-to-end transmission delay from the top root node to the bottom node, a file management method is used in chunks and files are exchanged from multiple sources.

As described above, the present invention implements a multicast function that supports multiple transmissions implemented as an overlay network on an IP network layer on a distributed content node, and performs multicast tree selection to improve such overlay multicast transmission performance. This paper proposes a topology, multicast tree selection method using node load information, file transfer per chunk to reduce end-to-end transmission delay, file exchange between clusters, and tree reselection due to node failure or overload.

This enables the application layer multicast function to efficiently support multiple transmissions using universal content nodes in the situation that IP multicast cannot provide universal network service, and the transmission quality of the network side, which is a disadvantage of the existing multicast. It can improve the problem, node load situation, group-wide transmission delay, and can reliably support multiple transmissions for network and node failures.

That is, in the present invention, a node receiving a chunk group can reduce the E2E delay rate by exchanging a chunk file with another node in the cluster, and simultaneously transmits a file from a main node to several local nodes. When the load is increased, there is an effect to improve the transmission quality degradation.

1 is a diagram illustrating an embodiment of an overlay multicast function and procedure according to the present invention;
2 is a functional block diagram of an embodiment of an overlay multicast system according to the present invention;
3 is a diagram illustrating an embodiment of an overlay multicast interworking protocol according to the present invention;
4 is an embodiment diagram illustrating an overlay multicast basic flow according to the present invention;
5 is a diagram for describing an overlay multicast file distribution and exchange method according to the present invention;
6 is a flowchart illustrating an embodiment of an overlay multicast file distribution and exchange process according to the present invention;
FIG. 7 is a diagram for explaining a multicast tree selection process using network context information and resource context information according to an embodiment of the present invention; FIG.
8 is a flowchart illustrating a multicast tree selection process using network context information and resource context information according to the present invention;
9 is an exemplary view illustrating the addition of a new node when overloading the road according to the present invention;
10 is an exemplary diagram for multicast tree reconstruction during node failure and overload according to the present invention;
11 is a functional block diagram of an entire system for overlay multicast according to another embodiment of the present invention;
12 is an explanatory diagram of an overlay multicast interworking protocol according to another embodiment of the present invention;
13 is a diagram illustrating an overlay multicast basic flow according to another embodiment of the present invention;
14 is a flowchart of an overlay multicast file distribution and exchange process according to another embodiment of the present invention;
15 is a flowchart illustrating a multicast tree selection process using network context information and resource context information according to another embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings, It can be easily carried out. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

And throughout the specification, when a part is referred to as being "connected" to another part, it includes not only "directly connected" but also "electrically connected" with another part in between. Also, when a component is referred to as " comprising "or" comprising ", it does not exclude other components unless specifically stated to the contrary .

In general, multicast technology is used to reduce the sender's data transmission load and network bandwidth usage in applications such as one-to-many streaming and one-to-many file transfer.

Although IP multicast technology is known as the most efficient way to support multicast by utilizing network resources, it is currently not widely used in real networks due to various obstacles.

Therefore, in the present invention, as an overlay network technology implemented on locally distributed content nodes without using IP multicast technology in a general-purpose content node to which the IP multicast function is not applied, for efficient one-to-many content transmission An application layer multicast system and a method thereof are provided.

The content node is a distributed node for distributing content to neighboring nodes close to the subscriber network (access network) to provide high quality content to customers without delay or interruption. The present invention provides an overlay method on an IP network layer of the content node. By implementing the multicast function, the transmission performance and stability of overlay multicast can be improved.

That is, the present invention relates to a method and apparatus for managing a group having the same interest on locally distributed content nodes to efficiently and high-quality delivering a message (file, content) to be delivered to the group.

The content node distributes large amounts of high-quality multimedia content close to the access network side, pre-positions or caches the content in advance, and delivers the content to customers in high quality without interruption or delay. ) Service node. The content node has a content distribution management structure for managing a large amount of content in a flexible, scalable and autonomous manner. Each content node has a unique node ID and has key values to be managed for each node ID. The range of key values that each node must manage is called the key range. If four nodes having a node ID of 6 bits have IDs of 13, 27, 43, and 57, respectively, the range of key values to be managed by each node is shown in Table 1 below.

Node ID Node Key Management Range 13 (001101) [13] to [26] 27 (011011) [27] to [42] 43 (101011) [43] to [56] 57 (111001) [57]-[63], [0]-[12]

In order to implement the multicast function in the application layer of the content node, the content node supports functions such as group creation, group join, group leave, group release, group message transmission, and the like for groups having the same interest.

To create a group, the node requesting creation creates a unique GroupID for a group with the same interest using a hash function (e.g. SHA256 ()), and creates a GroupID value (k ) Is selected as a root node of the group, and a group creation request is made to the selected root node. The root node that is requested to create a group manages the member information of the group through a group DB (Database).

Management information of the group DB is shown in Table 2 below.

Management topics Value (example) Group ID 28 (011100) Member 1 (Node ID) 12 (001100) Member 2 (Node ID) 38 (100110)

A node newly joining or leaving the group requests the group's root node to join or leave the group. The root node adds the node to the group DB or deletes the node from the group DB to join or leave the group.

The node that wants to send a message (file) to the group requests the root node to send the message. The root node stores the message in a multicast DB and gives the message a content delivery sequence number (sequential number of content to be sent to a particular multicast group or user). By assigning a content delivery sequence number to a message, messages to be transmitted in a group can be transmitted in chronological order.

Management information of the multicast DB managed by the root node is shown in Table 3 below.

Management topics Value (example) Group ID 28 (011100) Content Delivery Sequence Number 2 Sender ID (Node ID) 6 (000110) File ID 33 (100001)

1 is a diagram illustrating an embodiment of an overlay multicast function and procedure according to the present invention.

As shown in FIG. 1, in order to implement the overlay multicast function, the overlay multicast system includes functions of group creation, group join / leave, multicast message transmission, and group release.

In order to implement such a function, the overlay multicast system includes several functional entities, and the role of each functional block is illustrated in FIG. 2.

FIG. 2 is an exemplary functional block diagram of an overlay multicast system according to the present invention. The main functions of each block are as defined in the overlay multicast system functional block of Table 4 below.

Function block main function Service Routing Block (SR FE: Service Routing Functional Entity) Provides service routing within the overlay network
Receive service request, target service analysis (interpretation), forward request to target destination Identity Management (IDM) Block (FE) Manages identity authentication and permissions
Manage user profiles
Manages multicast group information Service Policy Decision (SPD: Service Policy
Decision Block (FE) Set the optimal multicast route using network status and resource information
Monitor network resource status and gather information
Applying Quality of Service Service Discovery and Negotiation (SDN) Block (FE) Manage content meta information (content title, size, supplier, rating, content location, etc.) Content Delivery (CD) Block (FE) Receive content, integrate content
Content Delivery: Delivers content to service requester from content aggregation through unicast or multicast. Overlay Management Block (FE) Status monitoring and information collection of service nodes Context Information Management (CIM) Block (FE) Collect and manage network, service, and node context information

3 is a diagram illustrating an embodiment of an overlay multicast interworking protocol according to the present invention, and a main function thereof is as defined in the overlay multicast interworking protocol of Table 5 below.

protocol Explanation Content location update The content delivery block (CD FE) requests the content location update to the service discovery and negotiation block (SDN FE). Content discovery The content delivery block (CD FE) requests the content location search to the service discovery and negotiation block (SDN FE). Multicast request The service routing block SR FE requests a multicast transmission to an end user or service to the content delivery block CD FE. Multicast Group Management The service routing block SR FE requests the creation, change and deletion of information on a multicast group to the identity management block IDM FE. Content storage & cache The Content Delivery Block (CD FE) manages its storage and cache. Multicast channel setup The content delivery block (CD FE) transmits a multicast channel creation request to the service policy decision block (SPD FE).
The service policy decision block (SPD FE) delivers a multicast channel setup command to the content delivery block (CD FE). Request context information The service routing block SR FE requests context information from the context information management block CIM FE. Request Resource Information The context information management block CIM FE requests resource information (node state, resource availability information) of the service node to the overlay management block OM FE.
The context information management block (CIM FE) manages the resource information of the node and provides resource context information when requested by another function block (FE) or service.

4 is a diagram illustrating an overlay multicast basic flow according to the present invention.

First, the context information management block CIM FE requests network context information from the service policy decision block SPD FE. The service policy decision block (SPD FE) then provides the network context information to the context information management block (CIM FE) by monitoring the network information (400). The context information management block CIM FE requests the resource context information from the overlay management block OM FE. The overlay management block OM FE then provides 401 the resource context information to the context information management block CIM FE by monitoring the node's dynamic information.

Thereafter, the service requester (terminal) requests the service routing block (SR FE) 1 to generate a multicast group (402).

Then, the service routing block (SR FE) 1 generates a group ID and selects a root node of the group (403).

The service routing block SR FE 1 requests the identity management block IDM FE to generate a multicast group (404).

The Identity Management Block (IDM FE) then creates a group DB (meaning a multicast group DB) (405).

Meanwhile, the service routing block (SR FE) 2 requests the service routing block (SR FE) 1 to join or leave the multicast group (406).

Then, the service routing block (SR FE) 1 forwards a request to join or leave the multicast group to the identity management block (IDM FE) (407).

The Identity Management Block (IDM FE) then updates (adds or deletes, etc.) the group DB (408).

The service routing block SR FE 1 requests context information from the context information management block CIM FE. The context information management block CIM FE 1 then provides the context information to the service routing block SR FE 1 (409).

Meanwhile, the content delivery block (CD FE) 1 requests the service policy decision block (SPD FE) to establish a multicast channel (410).

Then, the service policy determination block SPD FE forwards the multicast channel establishment request to the related content delivery blocks CD FE 1,2. Accordingly, the content delivery blocks CD FE 1,2 establish a multicast channel.

Meanwhile, the service routing block SR FE 1 requests the multicast to the content delivery block CD FE 1 (412).

Then, the content delivery block (CD FE) 1 requests the multicast DB update from the identity management block (IDM FE) (413).

The content delivery block (CD FE) 1 transmits the multicast content to the content delivery block (CD FE) 2 (414).

Thereafter, the service requester requests the service routing block SR FE 1 to release the multicast group (415).

Then, the service routing block (SR FE) 1 requests the multicast group release from the identity management block (IDM FE) (416).

Then, the identity management block (IDM FE) deletes the group DB (417).

5 is a diagram illustrating an overlay multicast file distribution and exchange method according to the present invention.

In overlay multicast, after the root node is selected, the transport tree is constructed. At this time, the transmission tree is mainly hierarchically composed of a root node, a main node, and a local node. The main node is a middle tier node based in major cities throughout the country, and the local node is the lowest neighboring node close to subscribers managed by the main node. For example, there may be about 10 main nodes in large cities such as Seoul, Busan, Daegu, and Daejeon, and about 20 regional nodes may be configured in about 20 Seoul, Busan, Daegu, and Daejeon.

The message transmission method is determined by the number of group members. When the number of group members is below the flooding threshold, the root node replicates the message and sends it one-to-one. However, when the number of group members is larger than the flooding threshold, the flooding method is used.

Flooding is transmitted from the root node to a hierarchical tree structure. In this case, when the file size to be transmitted is large, the receiving node receives a file and transmits the file to its subordinate node, thereby causing a long transmission delay. In addition, if the file size is large, the probability of transmission error is increased. When a file transmission error occurs, retransmission is attempted according to a specific value (number of retransmission attempts, for example, 3 times). Therefore, the delay time caused by this retransmission also becomes long. In order to improve this, in the present invention, when a file size is larger than a threshold size (for example, 200M), the file is divided into chunks of a predetermined size (for example, 2M) and stored and transmitted in chunk units. When the file is divided into chunks, a file sequence number is assigned, and the receiving node receives all the chunk files and reassembles the files according to the file sequence number.

By dividing the file into chunks in this manner, performance can be improved by performing parallel processing to simultaneously transmit files when transmitting files from one transmitting node to a plurality of lower nodes. In the present invention, transmission speed can be reduced by performing file transmission in parallel. When sending files from one node to multiple nodes in parallel, instead of sending the same chunk file to multiple nodes simultaneously for better performance, send different chunk files to reduce file I / O contention. do. In this case, there is a need for a method of appropriately grouping and transmitting the chunk files so as not to overlap when simultaneously transmitting the chunk files to the subordinate nodes. In the present invention, for example, a method of uniformly dividing the total number of chunks by the number of lower nodes is used.

That is, for example, when a movie is divided into 100 chunk files and transmitted, when the main node transmits to the local node, when the main node has 4 local nodes to transmit, the chunk numbers 1 to 25, 26 to 50, The chunk files are grouped into four groups, 51-75 and 76-100, and transmitted to each local node one-to-one. This ensures a balanced transfer of the entire chunk file as soon as possible while reducing input / output (I / O) contention while simultaneously transmitting non-overlapping chunk files to the four local nodes.

In addition, the present invention mainly considers that the overlay multicast tree structure is a structure in which the number of receiving nodes is concentrated at the lowest level, so that files are transmitted as quickly as possible from the intermediate node (main node) to the lowest node (local node) to reduce the transmission delay. In order to reduce the number of chunk files, the chunk file is distributed to each node in the cluster, and the chunk file is distributed once to each node in the cluster, and the chunk files are transferred to each node in the cluster. Use the exchange method.

For example, when transferring a file from one primary node to four local nodes (which are in a cluster), the chunk file is sent as 1 rather than the same chunk file four times in all four local nodes. Group chunks by 4 to transfer chunk files grouped by 1/4 to 4 local nodes one-to-one, and then use multiple nodes (sources) instead of one node for each local node in the cluster. The exchange delay can be reduced by exchanging chunk files.

At this time, when the chunk file is exchanged between local nodes in the same cluster, the chunk file location must be checked. The main node manages the chunk file information (content meta information) transmitted from each node to each local node in the cluster. Local nodes that have completed reception within the cluster update the reception completion to their local node and read the location of other chunk files from the main node to check the chunk file and the local node that they should receive, and then cluster. By exchanging multiple files with each local node in the network, the overall transfer latency can be reduced.

In the above-described embodiment, when the time when 100 chunk files are simultaneously transmitted from one primary node to four local nodes is T, the chunk files grouped by 25 at one primary node are 4 After transmitting to the local nodes, the transmission time of each local node in the cluster receiving 25 chunk files simultaneously from three local nodes except itself is reduced to T / 2.

In other words, as the number of nodes in the same cluster increases, the transmission speed increases compared with the conventional scheme. In other words, when a chunk file is exchanged between nodes, instead of receiving a file from one source node, the file is received from N-1 multiple sources except itself, so the transfer rate increases in proportion to the number of nodes N in the cluster. It can be seen that.

In the present invention, the local node receiving the chunk file group can reduce the E2E transmission delay by exchanging the chunk file with each local node in the cluster, and the transmission quality caused by the increased load when the primary node simultaneously transmits to multiple regional nodes. The degradation can be improved.

For example, if a movie file is divided into 100 chunk files and sent, the intermediate node (main node) sends the final node (local node) when the intermediate node has 4 local nodes to send. The chunk files are grouped as 26 ~ 50, 51 ~ 75, 76 ~ 100 and transmitted one-to-one to four local nodes. And four local nodes in a cluster can receive all the chunk files in a short time by exchanging chunk files with each other. Since four local nodes in a cluster can receive files from three other nodes at the same time, they have three times faster transfer speeds than files from one parent node. In this way, when a chunk file is exchanged in the same cluster, the file can be received more quickly by simultaneously receiving the file from multiple sources. Also, when exchanging files with other nodes, it is important to receive all files at the same time without duplication. For this purpose, the upper node (intermediate node) manages the chunk file and node ID that have been transferred to each cluster.

6 is a flowchart illustrating an embodiment of an overlay multicast file distribution and exchange process according to the present invention.

First, a service requester requests a multicast to a service routing block SR FE 1 (601).

The service routing block SR FE 1 then forwards the multicast request to the content delivery block CD FE 1 (602).

Then, the content delivery block (CD FE) 1 updates the multicast DB (603).

The content delivery block (CD FE) 1 requests the multicast channel establishment to the service policy decision block (SPD FE) (604).

Then, the service policy decision block SPD FE is composed of the content delivery block CD FE 1, the content delivery block CD FE 2,. In step 605, all content delivery blocks (CD FEs) related to up to content delivery block (CD FE) N are requested. Accordingly, each content delivery block (CD FE) establishes a multicast channel.

Then, the content delivery block (CD FE) 1 performs multicast content transmission to the content delivery block (CD FE) 2 (606).

The content delivery block (CD FE) 1 updates the content metadata (607). That is, the chunked file location (transmitted node ID) is updated and managed to which node each of the chunked content files is delivered.

Meanwhile, the content delivery block (CD FE) 2 determines (assigns) which chunk file to distribute to which subnodes (608).

The content delivery block (CD FE) 2 performs multicast content transmission to the content delivery block (CD FE) 5 which is its subordinate node (609). At this time, the content delivery block (CD FE) 2 does not transmit all the chunk files to the content delivery block (CD FE) 5 but only the chunk files allocated and distributed to the content delivery block (CD FE) 5.

The content delivery block (CD FE) 2 updates the content metadata (610).

The content delivery block (CD FE) 2 transmits the allocated chunk file to all content delivery blocks (CD FEs) in the cluster, respectively (611).

The content delivery block (CD FE) 2 distributes and distributes all the chunk files and updates the location of the chunk file in the content metadata to which node each chunk file is transmitted (612).

Meanwhile, the content delivery blocks (CD FEs) belonging to the same cluster request the chunk file from neighbor nodes in the cluster and receive the entire chunk file until they receive all the chunk files (613). That is, content delivery blocks (CD FEs) in the same cluster exchange chunk files with each other using multiple nodes (sources). At this time, the content metadata is used to inquire the location (node ID) of each chunk file to be requested.

In addition, the present invention can improve the stability of the transmission delay due to the quality of the lower network and the load and failure of the node on the multicast tree when transmitting the multicast tree (see FIG. 7 described later).

7 is a diagram illustrating an embodiment of a multicast tree selection process using network context information and resource context information according to the present invention.

When constructing a multicast tree in the application layer, the existing scheme cannot know network information in the application layer, so the tree is simply composed of the hierarchical structure of the IP network from the root node. The present invention utilizes a network proxy server (which serves as a context information management block) that collects routing information of each node and load information of nodes so that network information can be utilized in an application layer.

The network proxy server collects routing information of all nodes and collects node status and node load information. For this purpose, the node and the network proxy server need periodic interworking with the state information.

When creating a multicast tree after creating a multicast group, the optimal transport tree is selected by considering the topology information of the network proxy server and the load information of the node for the optimization of the transmission path. It is possible to configure the optimal transmission path.

8 is a flowchart illustrating a multicast tree selection process using network context information and resource context information according to the present invention.

First, the context information management block CIM FE requests resource context information from the overlay management block OM FE. Then, the overlay management block OM FE transmits the resource context information to the context information management block CIM FE by monitoring the resource information (801).

Thereafter, the service requester requests a multicast to the service routing block SR FE 1 (802).

The service routing block SR FE 1 then queries the identity management block IDM FE for multicast group information (group ID, group name, member, etc.) (803).

The service routing block SR FE 1 requests the multicast to the content delivery block CD FE 1 (804).

Then, the content delivery block (CD FE) 1 updates the multicast DB using the multicast request information (805).

The content delivery block (CD FE) 1 requests the service policy decision block (SPD FE) to establish a multicast channel (806).

The service policy decision block SPD FE then inquires the network context information from the context information management block CIM FE (807).

The service policy decision block SPD FE queries resource context information from the context information management block CIM FE (808).

The service policy decision block (SPD FE) sets an optimal multicast channel based on the network context information and the resource context information (809).

The service policy decision block (SPD FE) requests the multicast channel setup to all the content delivery blocks (CD FEs) in the group up to the content delivery block (CD FE) 1, ..., the content delivery block (CD FE) N. (810) Accordingly, each content delivery block (CD FE) establishes a multicast channel.

The service policy decision block (SPD FE) responds to the content delivery block (CD FE) 1 that the multicast channel setup is completed (811).

Then, the content delivery block (CD FE) 1 transmits multicast content to the content delivery block (CD FE) 2 which is its subordinate node by using the multicast channel (812).

Then, the content delivery block (CD FE) 2 performs multicast content transmission to its subordinate node (813). As such, all the lowest nodes transmit multicast content.

9 is an exemplary view illustrating the addition of a new node when overloading the road according to the present invention, and FIG. 10 is an exemplary view illustrating reconfiguration of a multicast tree when a node failure and an overload are performed according to the present invention.

When a performance degradation due to a failure or overload of a specific node occurs, the conventional method uses a static tree reconstruction method that predetermines a replaceable node. However, the present invention provides a method for dynamically replacing a node in case of failure. When determining, the optimal tree is reconstructed using the topology information of the node and the load information of the node.

In addition, the existing scheme has constructed a tree that selects its own primary node when a new node is added according to the regional information of the network regardless of the load information of the node. However, the present invention provides a network state (traffic congestion, Considering link failure, etc.) and the load state of the upper node, a tree considering transmission quality is constructed.

Hereinafter, another embodiment according to the present invention will be described with reference to FIGS. 11 to 15. In this case, the specific embodiment may be the same in a range corresponding to the embodiment described above with reference to FIGS. 1 to 10. Therefore, only the technical summary will be described briefly here.

FIG. 11 is a functional block diagram of an entire system for overlay multicast according to another embodiment of the present invention, illustrating a correlation for protocol definition between functional blocks of a service overlay network.

Here, the functions of the service routing block, the content delivery block, the service policy determination block, the context information management block, and the like shown in FIG. 11 correspond to the functions and operations of the blocks shown in FIG. Since other blocks are not related to the present invention, detailed description thereof will be omitted herein.

Meanwhile, detailed protocol definitions required for each component (functional block) for overlay multicast for multi-content distribution are shown in FIG. 12.

12 is an explanatory diagram of an overlay multicast interworking protocol for an overlay multicast system and a method according to another embodiment of the present invention.

Definition of each detailed protocol shown in FIG. 12 is as follows.

1. Content Delivery Request Process (corresponds to the Content Delivery Request Protocol in FIG. 12)

A) The service routing block SR FE forwards the content delivery request from the end user or service to the content delivery block CD FE.

B) At this time, the service routing block SR FE delivers user information, service information (service type, content ID), terminal information, and QoS parameters to the content delivery block (CD FE).

2. Content Delivery Channel Setup Process (corresponds to the Content Delivery Channel Setup Protocol of FIG. 12)

A) If the content delivery block (CD FE) delivers content in a multicast manner, the content delivery block (CD FE) establishes a multicast channel between the content delivery blocks (CD FE) and between services or end users. It sets using the context information (network context information, resource context information).

B) That is, the content delivery block CD FE sets an optimal multicast channel using context information (network context information, resource context information).

C) The Content Delivery Block (CD FE) establishes a multicast transmission path (i.e., multicast channel) and provides network context information (e.g. routing information, when a particular node fails, joins or leaves a multicast group). The multicast channel is reconfigured using the bandwidth, traffic, etc., and resource context information (e.g., the node's central processing unit (CPU), memory, network interface card (NIC), and disk usage).

D) Parameters of content delivery channel setting protocol are as follows.

Delivery method: multicast, unicast

Multicast group ID: unique identifier of the multicast group

-Multicast group name: name of the multicast group

Multicast group member: ID of the member node to join or leave the multicast group.

Multicast group hierarchy information: Multicast group hierarchy information such as parent node or child node of a specific node

3. Context Information Request Process (corresponds to Context Information Request Protocol in FIG. 12)

A) The service routing block SR FE requests the context information to the context information management block CIM FE.

B) The context information management block (CIM FE) provides network context information and resource context information to the service routing block (SR FE).

C) The service routing block SR FE provides the context information to the content delivery block CD FE for delivering content using the context information (network context information, resource context information).

D) The parameters of the context information request protocol are as follows.

Network context information: topology, bandwidth, traffic, etc.

Resource context information: CPU, memory, NIC, and disk usage, current session count

4. Transport QoS requirement delivery process (corresponds to the Transport QoS Dispatch Protocol in FIG. 12)

A) Service Policy Decision Block (SPD FE) is obliged to establish an efficient content delivery path using relevant network context information such as routing information, bandwidth, traffic, topology, etc.

B) The service policy decision block (SPD FE) provides QoS information for delivering content according to an optimal path of an underlying network.

C) The parameters of the transmission QoS dispatch protocol are as follows.

Routing Path: The primary or alternate routing path of the subnetwork.

Bandwidth: required bandwidth of service

-Traffic Status: Traffic Status of Sub Network

CPU, memory, NIC, disk usage: node CPU, memory, NIC, and disk usage

Current session number: The number of concurrent connection sessions of the node.

5. Transport QoS application process (corresponds to the Transport QoS Enforcement Protocol of FIG. 12)

A) Service Policy Decision Block (SPD FE) provides a mechanism to monitor resource status by querying sub-network or service requirements.

B) The Service Policy Decision Block (SPD FE) monitors dynamic information (eg, CPU, memory, NIC, and disk usage, current session count, etc.) of NGSON nodes.

C) Service Policy Decision Block (SPD FE) provides interworking with QoS-related entities in the underlying network to provide end-to-end QoS, thereby providing a QOS enforcement mechanism related to transmission.

D) Parameters of transmission QoS information protocol are as follows.

Routing Path: The primary or alternate routing path of the subnetwork.

Bandwidth: required bandwidth of service

-Traffic Status: Traffic Status of Sub Network

CPU, memory, NIC, disk usage: node CPU, memory, NIC, and disk usage

Current session number: The number of concurrent connection sessions of the node.

6. Content delivery process (corresponds to the Content Delivery Service Protocol of FIG. 12)

A) If the content delivery block (CD FE) delivers the content in a multicast manner, the content delivery block (CD FE) manages a group DB for managing the multicast group members in its local storage.

B) Multicast group management includes creation, modification and deletion of multicast DB.

In other words, the content delivery block (CD FE) manages the multicast DB in its local storage.

D) Multicast DB is a DB that records multicast requests. It is a DB for storing multicast requests so that multicasts can be sent sequentially in a specific group. The multicast DB includes a multicast group ID, a multicast sequence number (content delivery sequence number), a request node ID (sender ID), and a content ID (file ID).

E) If content is delivered using multicast, multicast group management uses message types such as group creation, group join, group leave, ungroup, multicast request, multicast response, and multicast NAK.

F) The multicast group information includes a multicast message type, a group ID, a group name, the requestor information includes a request node ID, and the service information includes a service type, a content ID, and a sequence number.

G) The parameters of the content delivery service protocol are as follows.

Delivery method: multicast, unicast

-Multicast message type: create, join, leave, release, multicast request, reply, multicast NAK

Multicast Group ID: Unique identifier of the multicast group.

-Multicast group name: name of the multicast group

Multicast group member: ID of the member node to join or leave the multicast group.

Request Node ID: ID of the node requesting the service

Content Delivery Sequence Number: A sequence number of content to be sent to a specific multicast group or user.

Content ID: Generates unique identifier and name of the content as hashed value.

Sequence no: sequence number of the chunked content file

Here, the content discovery protocol, the content location update protocol, and the content storage and cache protocol illustrated in FIG. 12 are additional components of the present invention. A sequence number (sequence no).

13 is a diagram illustrating an overlay multicast basic flow according to another embodiment of the present invention.

First, the context information management block CIM FE requests network context information and resource context information from the service policy decision block SPD FE. The service policy decision block (SPD FE) then provides network context information and resource context information to the context information management block (CIM FE) by monitoring the underlying network and monitoring the dynamic information of the nodes (1301).

Thereafter, the service requester (terminal) requests the service routing block (SR FE) 1 to generate a multicast group (1302).

Then, the service routing block (SR FE) 1 generates a group ID and selects a root node (representative node) of the group (1303).

The service routing block SR FE 1 requests the content delivery block CD FE 1 to generate a multicast group (1304).

Then, the content delivery block (CD FE) 1 generates a group DB (meaning a multicast group DB) (1305).

On the other hand, the service routing block (SR FE) 2 requests the content delivery block (CD FE) 1 to join or leave the multicast group (1306).

Then, the content delivery block (CD FE) 1 updates (adds or deletes, etc.) the group DB (1307).

Meanwhile, the service routing block SR FE 1 requests context information from the context information management block CIM FE. The context information management block CIM FE then provides context information to the service routing block SR FE 1 (1308). Accordingly, service routing block (SR FE) 1 provides context information to content delivery block (CD FE) 1.

Meanwhile, the content delivery block (CD FE) 1 establishes a multicast channel using context information (1309).

The content delivery block (CD FE) 1 transmits a multicast channel establishment request to the content delivery block (CD FE) 2 (that is, the child node content delivery block) that is its child node (1310). Accordingly, the content delivery block (CD FE) 2 also establishes a multicast channel.

Meanwhile, the service routing block SR FE 1 requests the multicast content transmission from the content delivery block CD FE 1 (1311).

Then, the content delivery block (CD FE) 1 updates the multicast DB using the multicast request information (1312).

The content delivery block (CD FE) 1 transmits the multicast content to the content delivery block (CD FE) 2 (1313).

Thereafter, the service requester requests the multicast group release from the service routing block SR FE 1 (1314).

Then, the service routing block SR FE 1 requests the content delivery block CD FE 1 to release the multicast group (1315).

Then, the content delivery block (CD FE) 1 deletes the corresponding group DB (1316).

14 is a flowchart illustrating an overlay multicast file distribution and exchange process according to another embodiment of the present invention.

First, a service requester requests a multicast from a service routing block SR FE 1 (1401).

The service routing block SR FE 1 then delivers the multicast request to the content delivery block CD FE 1 (1402).

Then, the content delivery block (CD FE) 1 updates the multicast DB (1403).

The content delivery block (CD FE) 1 establishes a multicast channel by setting multicast hierarchy information (child node ID) (1404).

The content delivery block (CD FE) 1 requests the multicast channel setup from the content delivery block (CD FE) 2 which is its child node (1405). Accordingly, the content delivery block (CD FE) 2 also establishes a multicast channel.

The content delivery block (CD FE) 2 is a multicast channel to the content delivery blocks (CD FEs) (content delivery block (CD FE) 5, ..., content delivery block (CD FE) N) which are child nodes thereof. Request setup (1406). Accordingly, each content delivery block (CD FE 5, ..., CD FE N) establishes a multicast channel.

Then, the content delivery block (CD FE) 1 performs multicast content transmission to the content delivery block (CD FE) 2 (1407).

The content delivery block CD FE 1 updates the content metadata (1408). That is, the chunked file location (transmitted node ID) is updated and managed to which node each of the chunked content files is delivered.

Meanwhile, the content delivery block (CD FE) 2 determines (assigns) which chunk file to distribute to which subnodes (1409).

The content delivery block (CD FE) 2 performs multicast content transmission to the content delivery block (CD FE) 5 that is its subordinate node (1410). At this time, the content delivery block (CD FE) 2 does not transmit all the chunk files to the content delivery block (CD FE) 5 but only the chunk files allocated and distributed to the content delivery block (CD FE) 5.

The content delivery block (CD FE) 2 updates the content metadata (1411).

The content delivery block (CD FE) 2 then transmits 1412 the allocated chunk file to all content delivery blocks (CD FEs) in the cluster.

The content delivery block (CD FE) 2 distributes and distributes all the chunk files, and updates the location of the chunk file in the content metadata to which node each chunk file is transmitted (1413).

Meanwhile, the content delivery blocks (CD FEs) belonging to the same cluster request the chunk file from neighboring nodes in the cluster until they receive all the chunk files and receive the entire chunk file (1414). That is, content delivery blocks (CD FEs) in the same cluster exchange chunk files with each other using multiple nodes (sources). At this time, the content metadata is used to inquire the location (node ID) of each chunk file to be requested.

15 is a flowchart illustrating a multicast tree selection process using network context information and resource context information according to another embodiment of the present invention.

First, the service policy decision block (SPD FE) monitors QoS information (topology, bandwidth, traffic, etc.) of the underlying network. In addition, the service policy decision block (SPD FE) monitors the dynamic information of the NGSON nodes (node status, node CPU, memory, NIC, disk usage, etc.) (1501).

The context information management block CIM FE requests network context information and resource context information from the service policy decision block SPD FE. The service policy decision block SPD FE then provides the context information management block CIM FE with network context information and resource context information monitoring the underlying network and the node 1502.

Thereafter, the service requester requests a multicast from the service routing block SR FE 1 (1503).

The service routing block (SR FE) 1 then queries the content delivery block (CD FE) 1 for multicast group information (group ID, group name, member, etc.) (1504).

The service routing block SR FE 1 requests the multicast to the content delivery block CD FE 1 (1505).

Then, the content delivery block (CD FE) 1 updates the multicast DB using the multicast request information (1506).

The service routing block SR FE 1 requests network context information and resource context information from the context information management block CIM FE (1507).

The content delivery block (CD FE) 1 sets an optimal multicast channel based on the network context information and the resource context information (1508).

The content delivery block (CD FE) 1 requests the multicast channel setup from the content delivery block (CD FE) 2 which is its subordinate node (1509). Then, the content delivery block (CD FE) 2 requests a multicast channel setup from the content delivery block (CD FE) N, which is its lower node (1510). In this way, all nodes in the group are sequentially requested to establish a multicast channel. Content delivery blocks (CD FEs) receiving a multicast channel establishment request establish a multicast channel by configuring their subnodes.

The content delivery block (CD FE) 1 transmits the multicast content to the content delivery block (CD FE) 2 which is its subordinate node by using the established multicast channel (1511). Then, the content delivery block (CD FE) 2 transmits multicast content to the content delivery block (CD FE) N, which is its subordinate node (1512). As such, multicast content transmission is performed to the lowest node.

On the other hand, the overlay multicast method according to the present invention as described above is implemented in the form of program instructions that can be executed by various computer means may be recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. The medium may be a transmission medium such as an optical or metal line, a wave guide, or the like, including a carrier wave for transmitting a signal designating a program command, a data structure, or the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Various permutations, modifications and variations are possible without departing from the spirit of the invention.

Therefore, the scope of the present invention should not be construed as being limited to the embodiments described, but should be determined by the scope of the appended claims, as well as the appended claims.

Claims (18)

In an overlay multicast system,
A service policy decision block (SPD FE) for monitoring the lower network and the nodes and delivering the context information to the context information management block;
The context information management block (CIM FE) for transferring the context information received from the service policy decision block to a service routing block;
The service routing block (SR FE) for delivering the context information received from the context information management block to a content delivery block and for delivering a content delivery request to the content delivery block; And
The content delivery block (CD FE) for establishing a multicast channel using the context information received from the service routing block and delivering the corresponding content through the set multicast channel according to a content delivery request from the service routing block. )
Overlay multicast system comprising a.
The method of claim 1,
The context information,
An overlay multicast system comprising network context information and resource context information.
The method of claim 2,
The network context information includes any one or more of routing information, bandwidth, and traffic;
The resource context information includes any one or more of the central processing unit (CPU) usage, memory usage, network interface card (NIC) usage, disk usage of the node.
The method according to any one of claims 1 to 3,
The content delivery block,
And in the event of node failure or multicast group join or leave, further performing the function of resetting the multicast channel by obtaining context information from the service routing block.
The method of claim 4, wherein
The content delivery block,
A group DB (Database) for managing multicast group members; And
Multicast DB for managing multicast request information
Overlay multicast system comprising a.
The method of claim 5, wherein
The content delivery block,
And a sequence number indicating a sequence number of the chunk file obtained by dividing the corresponding content in chunk units.
In the overlay multicast method,
Monitoring, by the service policy determination block, the lower network and the node, and delivering the context information to the context information management block;
Transmitting, by the context information management block, context information received from the service policy determination block to a service routing block;
Delivering, by the service routing block, context information received from the context information management block to a content delivery block;
Establishing, by the content delivery block, a multicast channel using the context information received from the service routing block;
The service routing block forwarding a content delivery request to the content delivery block; And
Delivering, by the content delivery block, the corresponding content through the established multicast channel according to a content delivery request from the service routing block;
Overlay multicast method comprising a.
The method of claim 7, wherein
Resetting the multicast channel by acquiring context information from the service routing block in the event of a node failure or joining or leaving a multicast group;
Overlay multicast method further comprising.
9. The method according to claim 7 or 8,
The context information,
An overlay multicast method comprising network context information and resource context information.
9. The method according to claim 7 or 8,
The content delivery block,
Manage multicast group members using group DB, manage multicast request information using multicast DB,
And managing a sequence number indicating a sequence number of the chunk file obtained by dividing the corresponding content in chunk units.
In an overlay multicast system,
A context information management block (CIM FE) for requesting and obtaining context information from a service policy decision block;
Service routing for acquiring context information from the context information management block and delivering it to a content delivery block, selecting a group ID and a root node according to a multicast group creation request, and requesting the content delivery block to generate a multicast group. Block SR FE;
Create a group DB according to a multicast group creation request from the service routing block, update the group DB according to a multicast group join or leave request, and set up a multicast channel using context information from the service routing block. And updating the multicast DB using the multicast request information according to the multicast content transmission request from the service routing block, and transmitting the content to a lower content delivery block through the set multicast channel. (CD FE)
Overlay multicast system comprising a.
The method of claim 11,
The content delivery block,
And in the event of node failure or multicast group join or leave, further performing the function of resetting the multicast channel by obtaining context information from the service routing block.
13. The method according to claim 11 or 12,
The content delivery block,
And deleting the group DB according to a multicast group release request from the service routing block.
13. The method according to claim 11 or 12,
The content delivery block,
The content is divided in chunks and divided and distributed to content delivery blocks in the same cluster, and the content metadata indicating the location of each chunk file (each node to which each chunk file is delivered) is updated and managed.
The content delivery blocks in the same cluster identify the location of the chunk file using the content metadata and exchange the chunk files with each other using multiple nodes.
In the overlay multicast method,
Obtaining, by the context information management block, context information from the service policy decision block;
The service routing block selecting a group ID and a root node according to the multicast group creation request and requesting the content delivery block to generate the multicast group;
Generating, by the content delivery block, a group DB and updating the group DB according to a request for joining or leaving a multicast group from the service routing block;
Obtaining, by the service routing block, context information from the context information management block and delivering the context information to the content delivery block;
Establishing, by the content delivery block, a multicast channel using context information; And
Updating, by the content delivery block, a multicast DB using multicast request information according to a multicast content transmission request from the service routing block, and transmitting content to a lower content delivery block through the set multicast channel;
Overlay multicast method comprising a.
The method of claim 15,
Resetting the multicast channel by acquiring context information from the service routing block in the event of a node failure or joining or leaving a multicast group;
Overlay multicast method further comprising.
17. The method according to claim 15 or 16,
Deleting, by the content delivery block, the group DB according to a multicast group release request from the service routing block;
Overlay multicast method further comprising.
17. The method according to claim 15 or 16,
The content delivery block,
The content is divided in chunks and divided and distributed to content delivery blocks in the same cluster, and the content metadata indicating the location of each chunk file (each node to which each chunk file is delivered) is updated and managed.
And content delivery blocks within the same cluster identify the location of the chunk file using the content metadata to exchange chunk files with each other using multiple nodes.

Images