CN101938508B - Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system - Google Patents
Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system Download PDFInfo
- Publication number
- CN101938508B CN101938508B CN2009101487269A CN200910148726A CN101938508B CN 101938508 B CN101938508 B CN 101938508B CN 2009101487269 A CN2009101487269 A CN 2009101487269A CN 200910148726 A CN200910148726 A CN 200910148726A CN 101938508 B CN101938508 B CN 101938508B
- Authority
- CN
- China
- Prior art keywords
- routing server
- cache routing
- peer client
- peer
- cache
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a method and a system for shortening time delay in a peer-to-peer network streaming media live broadcast system. The system is provided with a media source node and a cache routing server, wherein the media source node is used for the media resource management of the whole system; the cache routing server is used for media resource management, routing calculation and peer client end management in a local domain; the cache routing server is used for selecting a proper peer client end from other managed peer client ends and establishing streaming media data interaction with a request peer client end; when needed media streaming data does not exist in the peer client end managed by the cache routing server, the cache routing server is used for requesting needed streaming media data information from other cache routing servers and transmitting the needed streaming media data information to the request peer client end; and when the cache routing server is failed for obtaining the needed streaming media data from other cache routing servers, the cache routing server is used for obtaining the needed data from the media source node and transmitting the needed data to the request peer client end.
Description
Technical field
The present invention relates to the Technology of Network Streaming Media field, refer to especially the method and system that time-delay reduces in a kind of peer-to-peer network (P2P, peer-to-peer) flow medium live system.
Background technology
Along with the development of the Internet, utilizing network to carry out streaming media playing becomes a focus.Streaming Media refers to the multimedia of (Internet)/in-house network (Intranet) use stream transmission technology in the Internet, and Streaming Media does not need to download whole file before broadcast, can watch while downloading.Its principle be to the multimedia file of user transmission according to broadcast time order be divided into different segments, then successively these segments are issued the user, play continuously these segments and just formed continuous sound and image.Traditional network flow-medium Play System is generally based on client/server (C/S) pattern.But because streaming media playing need to consume a large amount of network bandwidths and server resource, therefore, this pattern has seriously limited the development of network flow-medium business.Network video stream take an occupied bandwidth as 300kbps is as example, and when having 1000 users to watch simultaneously, the bandwidth that needs is 300Mbps.Such performance requirement is what not satisfy for most of servers.If support several ten thousand even the user of hundreds of thousands, it almost is impossible adopting this C/S model.Certainly, increasing number of servers and improving bandwidth is a settling mode, but this mode needs huge cost, so we need to address this problem by new technological means.Internet protocol (IP) multicast is to solve a kind of technology that the network flow-medium multiplexing consumes massive band width.Utilize multicast, a piece of data can be sent to simultaneously one group of specific user, thereby avoided the transmission that repeats of data, can not cause the waste of the network bandwidth, be fit to very much Network Transmission Streaming.But the IP multicast also has the problem of self.Because Internet is a heterogeneous network, hardware platform varies, and interconnect equipment also there are differences in the support to multicast, and therefore, there is the difficulty of deployment in actual applications in the IP multicast.
Peer-to-peer network is emerging network technology recent years, and with respect to traditional C/S model, the characteristics of a highly significant of P2P pattern are exactly that node need not Dependency Set Chinese style server resource, and each node can directly communicate.Each node has identical status, both can request service, also can provide service, and play the part of simultaneously the dual role of server and client computer in the C/S model, even can also have the function of router and high-speed cache.Under the P2P pattern, because each node has the function of client-server, like this, each node is also sending data flow to other nodes in other node receiving data stream.Utilize this characteristics, the P2P network technology can use existing computational resource and the finite element network bandwidth realizes the large scale network live streaming media.
The problem that at first needs to solve in the P2P stream media system is to allow the node (peer) of new adding obtain program data from which or which node, and allows this peer node which node to provide data, services for.Also be the problem that the peer node topology is searched in essence.Most work is that the peer node organization is become multicast tree in existing research, the data that the nodes sharing on every multicast tree flows out from channel of source server, and each node only participates in a multicast tree.This Data dissemination mode is closed the genus one-to-many pattern of fastening ISP and service consumer number.Owing to existing based on the Data dissemination of single multicast tree to can not get such as the leaf node bandwidth resources that leaving of effective utilization, intermediate node or failure behaviour are large on the impact of child node, service quality (QoS) is difficult to the problems such as guarantees, people so also proposed based on the Data dissemination of many multicast tree with based on the schemes such as Data dissemination of random topology.In the Data dissemination scheme based on many multicast tree, data flow is divided into many subflows at the source server end, every subflow is transmitted with an independent multicast tree, and each peer node can join according to the bandwidth resources situation of oneself in single or many multicast tree.Although the multicast that many trees scheme proposes has solved a lot of problems of single tree-model to a certain extent, owing to still still having based on the basic multicast structure of tree topology, it much is difficult to the problem that overcomes.
Multicast tree is divided into the complexity that multicast has increased the multicast structure greatly:
The degree of depth of tree is restive equally, and particularly after the degree of depth of tree reached certain level, the delay of different multicast tree may be unbalanced in the multicast, thereby it is satisfied to cause program quality to be difficult to;
Multicast tree is difficult to balance when withdrawing from when adding frequently in the face of a large amount of line nodes, certain a multicast tree uneven even can cause the collapse of whole multicast in the multicast;
Key node accident in the multicast withdraws from or service ability can not satisfy the time, and the service quality of its whole descendant nodes is difficult to guarantee equally;
The adding escape mechanism of node is more complicated, and the joining day is longer;
The maintenance costs of multicast is too high, and expense is different from the tree model along with the destructuring multicast structure of the exponential Cool of the increasing progressively Streaming/DONET of the number of the multicast tree employing that is representative based on the Gossip agreement.Cool Streaming/DONET provides the destructuring multicast model of an approximate diagram shape, the flow direction of data is no longer single and more changeable in this model, need to introduce special deblocking and scheduling mechanism to adapt to this structure, it has following advantage:
Have better expandability on the structure, can support more user;
Can solve preferably the service quality decline problem that heterogeneous nodes and asymmetric node bring, more suit to use at Internet;
Stability better, it is less that node improper withdrawed from whole multicast structure influence, even the online multicast structure of a large number of users also is easy to balance;
Postpone less, after number of users reached certain level, for the tree shaped model of equal number of users, the jumping figure that data arrival fringe node is passed by was less, and the delay nature is lower.
But Cool Streaming/DONET model self also has some shortcomings:
The degree of freedom on a node basis is too high, for operator without controllability;
Do not combine with the bottom physical topology;
DONet has obtained network efficiency in the reliability that has guaranteed system, but has sacrificed time-delay.Subject matter among the DONet is the startup time-delay of system and the adaptability of network isomery bandwidth.
Summary of the invention
In view of this, the object of the invention is to propose the method and system that time-delay reduces in a kind of P2P flow medium live system, strengthen Overlay and bottom physical network matching properties that stream media system makes up, shorten the time-delay of playing, the cache routing server significantly reduces the startup time-delay that node is play by pushing media resource to the peer client.
Based on the method that time-delay in above-mentioned purpose a kind of Peer-to-Peer Network P2P flow medium live system provided by the invention reduces, in system, dispose medium source nodes, be used for the managing medium resource of whole system; And more than one cache routing server, the managing medium resource, the route that are used in the local field are calculated and the peer client-side management; And comprise:
After the stream medium data of receiving the peer client received request, the cache routing server was selected suitable peer client in other peer clients of managing, and the stream medium data between foundation and the request peer client is mutual;
When the media stream data of asking the peer client to need did not exist in the peer of this cache routing server admin client, the cache routing server was pushed to request peer client to the required stream medium data information of other cache routing server request;
When the cache routing server obtained required stream medium data failure from other cache routing server, the cache routing server obtained desired data from medium source nodes, is pushed to request peer client.
Optionally, the method also comprises: medium source nodes sends the multidate information of self-administered cache routing server list to each cache routing server, and timed sending inquires that heartbeat message that whether this cache routing server survive is to corresponding cache routing server;
Cache routing server response self heartbeat message comprises bandwidth and the ability value coded message of self to medium source nodes in heartbeat message;
The cache routing server list that medium source nodes is safeguarded according to the information updating of responding.
Optionally, the method also comprises: the cache routing server upgrades in local peer client side list for the adding of the peer client of managing and the behavior of leaving;
The cache routing server is processed the peer client side list of managing and this cache routing server according to the behavior of peer client result sends to medium source nodes.
Optionally, the method is fashionable when there being new peer client to add, and the method further comprises: the principle that medium source nodes is preferential according to the region is assigned to corresponding cache routing server with new peer client;
New peer client adds to the cache routing server request, the cache routing server is according to the channel of this new peer client-requested, in local cache, adopt time coordinate to retrieve needed media resource, if there is current in progress media resource, the stream medium data fragment is pushed to new peer client;
If required media resource does not exist, then to medium source nodes request media resource, the cache routing server obtains the stream medium data frag info from medium source nodes by the mode of drawing;
After the cache routing server gets access to media resource, be pushed to the peer client of request, upgrade simultaneously the media resource buffer memory of safeguarding.
Optionally, the method also comprises: between each cache routing server the peer client side list of managing is undertaken by heartbeat message alternately.
Optionally, the described cache routing server of the method is selected suitable peer client in other peer clients of managing, the stream medium data between foundation and the request peer client also comprises alternately: the peer client is calculated the ability value of the current disposal ability that shows this peer client according to current bandwidth and CPU usage; The peer client is attached to own ability value and BM block message and sends to the cache routing server in the heartbeat message;
When the stream medium data of receiving the peer client receives request, the cache routing server is according to the described ability value and the BM block message that obtain, adopt Hierarchical Matching Algorithm to carry out the suitable peer client of route calculative determination, and route result of calculation is sent to described request peer client.
Optionally, the described route computational process of the method comprises:
A. according to the order incremental of the client ip network segment of living in IP index of definition to every section network of database, for the multilayer network segment, successively set up index successively until top from minimum IP network section;
B. the index IP that sets up with client ip and steps A compares coupling, judges near distance, according to the division of existing IP four hop count values, from high section successively coupling until low section;
C. according to network layer match index in turn, from top successively deeply, until lowermost layer, the final selected zone of dividing, and this IP carried out in addition network area sign;
D. preferentially mate according to the area flag position;
E. with matching result ordering, according to ranking results, one by one the BM piece is compared, therefrom select to have the forward peer client of required BM piece and IP matching addresses as route results.
Optionally, when the data that the peer client of the described cache routing server admin of the method needs do not exist in the peer of this cache routing server admin client, further comprise to the process of other cache routing server request data:
If the cache routing server load of requested date surpasses predefined threshold value, then this requested cache routing server is selected a peer client from the peer client of managing, and returns to request peer client as route results.
Optionally, when described other cache routing server of the method and the peer client of managing thereof do not have desired data or do not provide data, services, the cache routing device is also judged under the peer client: if the current media source node then provides data, services by medium source nodes under load threshold; If the current media source node surpasses load threshold, then denial of service.
Optionally, the described cache routing server of the method comprises from the situation that other cache routing server obtains required stream medium data failure: other cache routing server and the peer client of managing thereof do not have desired data or data, services are not provided.
Based on above-mentioned purpose, the present invention also provides the realization system that time-delay reduces in a kind of peer-to-peer network (P2P) flow medium live system, it is characterized in that, comprising:
Medium source nodes is for the managing medium resource of whole system;
More than one cache routing server, the managing medium resource, the route that are used in the local field are calculated and node administration; Described cache routing server is selected suitable peer client in other peer clients that it is managed, the stream medium data between foundation and the request peer client is mutual; When the media stream data of asking the peer client to need did not exist in the peer of this cache routing server admin client, the cache routing server was pushed to request peer client to the required stream medium data information of other cache routing server request; When the cache routing server obtained required stream medium data failure from other cache routing server, the cache routing server obtained desired data from medium source nodes, is pushed to request peer client.
Optionally, the described medium source nodes of this system further comprises:
The district management unit according to number of users and distributional region, creates and management domain, according to cache routing server in the selection of device resource characteristics and the specified domain, obtains the resource statistics information in each territory; Determine the merge and split in territory according to the quantity of member in the territory;
The node administration unit, by with OSS user authentication module interface, peer ID, the payment information of authentication peer client are guaranteed its legitimacy; According to terminal characteristics such as address, content, specify it to add the domain of dependence; Server, the peer resource of management the whole network.
Route computing unit carries out Preliminary screening to the whole network resource data, regularly to the overseas routing iinformation of cache routing server push; According to cache routing server route requests, inquire about the whole network resource statistics information of its storage, select to the relevant routing iinformation of other cache routing server request;
The network management interface unit has the platform management module interface in OAMAgent Agent and the OSS, realizes management, maintenance software upgrading.
Optionally, this system cache routing server comprises:
The node resource administrative unit, the resource information of peer client, overseas resource information in storage and the management domain;
The node administration unit, withdraws from the adding of peer client in the management domain; By with the territory in the communication interface of peer client, regularly with the communication of peer client, monitor the peer client state.
Route computing unit with the medium source nodes communication, obtains up-to-date resource routing information; According to the feature of peer client, to the primary Calculation of resource data, regularly to peer client push routing iinformation; To take into account the territory internal burden during calculating balanced; The route requests of peer client represents the overseas route that the peer client please be looked for novelty to medium source nodes in response and the Acknowledge;
The network management interface unit has the platform management module interface in webmaster Agent and the OSS, realizes management, maintenance, the software upgrading of cache routing server.
Can find out from above, the method and system that time-delay reduces in the P2P flow medium live system provided by the invention, by in system, disposing medium source nodes and buffer memory routing server, form three grades of P2P media network structures by peer client, cache routing server, source of media.In network operation process, by source of media, the information interaction of cache routing server and peer client, tissue forms network topology structure, makes up the Overlay network.The cache routing server adds fashionable in new peer client, by the propelling movement mode, provide the data slot of current broadcast for the peer client, reduces the startup time-delay of peer client terminal playing.When the cache routing server calculates route, according to the layered matching process of IP address, the exchange of data is remained between the node of bottom physical network coupling as far as possible, reduce consumption and the load of node bottom-layer network bandwidth.By the information exchange between different cache routing servers, when the peer client is obtained less than valid data, can cross over regional access to data in the one's respective area.Use this method by the propelling movement of cache routing server, reduce live startup time-delay.By strengthening nerve of a covering and the physical network matching of P2P stream media system, improve media data and obtain efficient, reduce the live play time-delay.
Specifically comprise following beneficial effect:
1) powerful controllability, because cache routing server and medium source nodes all are to dispose management by operator, the behavior of peer client by cache routing server and medium source nodes control, therefore, can obtain the control to the P2P stream media network fully.
2) significantly reduce to play the startup time-delay, the cache routing server initiatively is pushed to newly added node with the media data of current broadcast, new node can be play immediately, and node carries out the P2P network in playing process topology generates, and has significantly reduced the startup time-delay of playing.
3) significantly reduce broadcast time-delay, carry out unified routing management by the cache routing server, reduced the blindness of partner's search.By IP address hierarchical index so that in the bottom physical network contiguous node intercourse the video stream media data; Be maintained in the bottom physical network relevant informations such as host performance (only having considered processing speed and storage here), bandwidth availability ratio of contiguous ordinary node by the cache routing server, and according to these relevant informations node managed.Reduced by this method the Bonding pressure of medium source nodes and buffer memory routing server and end-to-end data transfer delay, and the reduction of these two parameters can greatly reduce the time-delay in the playing process.
4) simple for structure, adopt the way of DoNet between node, the transmission of node administration and stream medium data has separately been carried out, come the distribution flow media data according to the flow direction that data transmit, dispose again the cache routing server and come management node, do not needed to safeguard very complicated data structure.
5) good extensibility, do not need to safeguard a large amount of data structures, each ordinary node is only preserved the information of other nodes of its needs, and the cache routing server only is in charge of and safeguard and its nodal information in a physical network, the increase that adds the video stream media network along with node, performance is better, strong adaptability.
Description of drawings
Fig. 1 is embodiment of the invention system configuration and node reciprocal process schematic diagram;
Fig. 2 is embodiment of the invention medium source nodes internal structure schematic diagram;
Fig. 3 is embodiment of the invention cache routing server internal structure schematic diagram.
Embodiment
For making the purpose, technical solutions and advantages of the present invention clearer, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in more detail.
The technical scheme of the method that time-delay reduces in the Peer-to-Peer Network P2P flow medium live system of the present invention mainly comprises:
In system, dispose medium source nodes, be used for the managing medium resource of whole system; And the part of server in the system (more than) is deployed as the cache routing server, the managing medium resource, the route that are used in the local field are calculated and the peer client-side management; Other ordinary nodes are as the peer client.
After the stream medium data of receiving the peer client received request, the cache routing server was selected suitable peer client in other peer clients of managing, and the stream medium data between foundation and the request peer client is mutual;
When the media stream data of asking the peer client to need did not exist in the peer of this cache routing server admin client, the cache routing server was pushed to request peer client to the required stream medium data information of other cache routing server request;
When the cache routing server obtained required stream medium data failure from other cache routing server, the cache routing server obtained desired data from medium source nodes, is pushed to request peer client.
For such scheme following explanation is arranged: it is mutual to carry out stream medium data between the peer client of cache routing server admin, this is the predominating path of video stream media exchanges data in the network, when calculating route, adopts the cache routing server method with IP address layering and matching, the exchange of data is remained between the node of bottom physical network coupling as far as possible, reduce consumption and the load of node bottom-layer network bandwidth.
When the data of the peer of cache routing server admin client needs did not exist in the peer of this cache routing server admin client, the cache routing server was just to other cache routing server request data message; When requested cache routing server does not provide data, services, manage the peer client from this requested cache routing server and to select, reply to request peer client, the peer client is set up data with it and is connected.
When not having peer client that the data that need and other territory do not have required data in the self-administered peer client of cache routing server, the cache routing server obtains the peer client that desired data directly sends to request from other cache routing server.
When other cache routing server and the peer client of managing thereof did not have desired data or do not provide data, services, the cache routing server obtained desired data from medium source nodes, was pushed to request peer client.
In the preferred embodiment of the present invention, medium source nodes and buffer memory routing server regularly carry out the two-way interactive of control information, mainly comprise: the process information that a) dynamic-change information of cache routing server admin peer client, and cache routing server is carried out the peer client.B) medium source nodes send self-administered cache routing server list multidate information to each cache routing server, and timed sending inquires that heartbeat message that whether this cache routing server survive is to corresponding cache routing server; After the cache routing server is received inquiry message, can respond self heartbeat message to medium source nodes, in heartbeat message, comprise bandwidth and the ability value coded message of self; The cache routing server list that medium source nodes is safeguarded according to the information updating of responding.
Also carry out two-way control information between peer client and the buffer memory routing server mutual: the cache routing server sends to the routing iinformation that chooses the peer client of request route; Each peer client is own available bandwidth, and the dynamic-change information such as disposal ability periodically send to the cache routing server.
In addition, the cache routing server upgrades in local peer client side list for the adding of the peer client of managing and the behavior of leaving;
The cache routing server is processed the peer client side list of managing and this cache routing server according to the behavior of peer client result sends to medium source nodes.
Between each cache routing server the peer client ip tabulation of managing is undertaken by heartbeat message alternately.
In the preferred embodiment of the present invention, the cache routing server obtains the stream medium data frag info from medium source nodes by the mode of drawing, in the situation that the medium source nodes load is heavier, forbid that other nodes directly obtain data from medium source nodes, only allow the cache routing server directly to obtain data from medium source nodes.Other nodes described here not only comprise the peer client, also comprise other cache routing servers etc., that is to say that medium source nodes should have the linking number restriction, when linking number transfinites, refuses the connection of other any nodes.
The peer client exchange stream medium data information that cache routing server and it are managed, when the peer of cache routing server admin number clients was fewer, it provided the part available bandwidth to other peer client service; And when the peer number clients was increased to a certain degree, then this cache routing server only can provide route service and initial propelling data service, and data exchange service is not provided; The current plays clip of cache routing server active push is given the new peer client that adds.
Below in conjunction with the drawings and specific embodiments, technical solution of the present invention is elaborated.
Referring to system configuration shown in Figure 1, the realization system that time-delay reduces in the P2P flow medium live system that the present invention proposes comprises:
Medium source nodes 10 is for the managing medium resource of whole system; Cache routing server 20, the managing medium resource, the route that are used in the local field are calculated and node (comprising peer client 30 and buffer memory routing server 20) management; And peer client 30.
Wherein, medium source nodes 10 is centralized-control centers of resource, requires equipment can meet the requirement of telecom operation, has the carrier-class reliability and stability, is set up by operator.
Referring to shown in Figure 2, medium source nodes 10 mainly comprises:
Network management interface unit 204 has the platform management module interface in OAMAgent (Agent) and the OSS, realizes management, maintenance software upgrading.
Cache routing server 20 is district management centers, is in charge of the regional interior nodes resource of (mainly comprising the peer client).
Referring to shown in Figure 3, mainly the comprising of cache routing server 20:
Node resource administrative unit 301, the resource information (such as user ID, contents fragment etc.) of peer client 30, overseas resource information (such as user ID, contents fragment etc.) in storage and the management domain.
Network management interface unit 304 has the platform management module interface in webmaster Agent and the OSS, realizes management, maintenance, the software upgrading of cache routing server.
In addition, in medium source nodes 10 and buffer memory routing server 20, all have network interface (Net Interface), be used for being connected with other nodes (comprising medium source nodes, cache routing server, peer client etc.) of network.
Referring to shown in Figure 1, an embodiment of time-delay minishing method comprises the steps: in the P2P flow medium live system of the present invention
1) information exchange and the maintenance between medium source nodes and the buffer memory routing server.
Step 101, medium source nodes send the multidate information of self-administered cache routing server list to each cache routing server, and timed sending inquires that heartbeat message that whether this cache routing server survive is to corresponding cache routing server.Wherein, comprise at least in the routing server tabulation: cache routing server address information, and state information corresponding to these cache routing servers.
Step 102, cache routing server response self heartbeat message is to medium source nodes.The bandwidth and the ability value coded message that in heartbeat message, comprise self.The cache routing server list that medium source nodes is safeguarded according to the information updating of responding.
Step 103, the cache routing server is processed the peer client side list of managing and cache routing server according to the behavior of peer client result sends to medium source nodes.For the adding of peer client with leave behavior cache routing server and in the peer client side list of managing, upgrade.Wherein, the peer client side list mainly comprises: the information such as the numbering of peer client, IP address, state.
2) fashionable when there being new peer client to add, new peer client adds and just means request media resource, execution in step 104-107 generally speaking.
Step 104, the principle that medium source nodes is preferential according to the region is assigned to corresponding cache routing server with new peer client.
Step 105, new peer client adds to the cache routing server request, the cache routing server is according to the channel of its request, in buffer memory, adopt time coordinate to retrieve needed media resource, if there is current in progress media resource, 60 second flow media data fragments are pushed to new adding peer client.After this peer client is obtained data, can play immediately, shorten and start time-delay; Do not exist if be fit to media resource, to medium source nodes request media resource.
Step 106, the cache routing server obtains the stream medium data frag info from medium source nodes by the mode of drawing.Can be too not congested for the load of controlling medium source nodes, can load threshold be set for medium source nodes, and only allow the cache routing server directly to obtain data from medium source nodes.Wherein, described load threshold can arrange flexibly, and for example: linking number that can the restricting media source node is 1000, perhaps limits cpu busy percentage to 80%, and perhaps when limiting bandwidth utilance to 90%, medium source nodes is not being accepted new connection request.
Step 107 after the cache routing server gets access to media resource, is pushed to the peer client of request, upgrades simultaneously the media resource buffer memory of safeguarding.
In addition, when the peer of cache routing server admin number clients is fewer, this cache routing server is to the load management less of control information, and then it can provide the part available bandwidth to other peers video stream media data, serves as the Peer client; And when the peer number clients is increased to a certain degree, this cache routing server is relatively large to the load management of control information, then this cache routing server only provides for other peers provide route and initial flow media data, and data exchanging function is not provided, guaranteed that the load of cache routing server remains on to a certain degree.
3) in step 108-110, can also manage the information exchange of peer client side list between cache routing server and other cache routing servers.
Step 108 is undertaken the peer client side list of managing by heartbeat message between each cache routing server alternately.
Step 109, single Peer client can calculate an ability value according to current bandwidth and CPU usage, and this value shows the current disposal ability of this Peer client.Single peer client sends to the cache routing server to own ability value and BM block message by being attached in the heartbeat message.
Step 110, when the stream medium data of receiving the peer client received request, the cache routing server was according to the described ability value and the BM block message that obtain, and the cache routing server adopts Hierarchical Matching Algorithm to carry out route and calculates.Carry out Optimum Matching according to the IP address, coupling is carried out according to following steps:
A. according to the order incremental of the client ip network segment of living in IP index of definition to every section network of database.For the multilayer network segment, successively set up successively index IP until top from minimum IP network section.Wherein, cache routing server inside can utilize described database to store these index IP and BM block message, with the fast and easy reading writing information.
B. the index IP that sets up with client ip and steps A compares coupling, judges near distance.According to the division of existing IP four hop count values, from high section successively coupling until low section.
C. according to network layer match index in turn, from top successively deeply, until lowermost layer, the final selected zone of dividing, and this IP carried out in addition network area sign.
D. preferentially mate according to the area flag position.
E. with matching result ordering, according to ranking results, one by one the BM piece is compared, therefrom select to have the forward peer client of required BM piece and IP matching addresses as route results.
When 4) data of the peer client of cache routing server admin needs do not exist in the peer of this cache routing server admin client, can also in other cache routing servers and its peer client of managing even medium source nodes, seek by following steps 111-114.
Step 111 is not when the data that the peer client of cache routing server admin needs exist in the peer of this cache routing server admin client, to other cache routing server request data.Wherein, consideration needs the loading condition of the cache routing server of swap data, if the cache routing server load of requested date is excessive, surpasses predefined threshold value, and then this requested cache routing server just can not provide data, services.This requested cache routing server can be selected a peer client (routing iinformation of response) from the peer client of managing, return to request peer client as route results.Wherein, described threshold value can arrange flexibly, and for example: the linking number that can limit the cache routing server is 1000, perhaps limits cpu busy percentage to 80%, and perhaps when limiting bandwidth utilance to 90%, the cache routing server is not being accepted new connection request.
Step 112, the cache routing server is with the route the calculated peer client response to the request route.
Step 113 after the peer client is received route results, is set up data according to route results with peer client or cache routing server and is connected, the exchange current media data, and this is the predominating path of video stream media exchanges data in the network.
Step 114, if when the cache routing server can not find the route that needs, medium source nodes can provide data, services by medium source nodes under load threshold simultaneously.Just obtain data from medium source nodes, be pushed to request peer client, if the current media source node surpasses the load threshold of self, then denial of service.
Need to prove above step 1)-4) and represent execution sequence, can directly enter corresponding step according to different situations in the practical application.
Above-described specific embodiment is specific embodiments of the invention only, is not limited to the present invention, and is within the spirit and principles in the present invention all, any modification of making, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (13)
1. the method that time-delay reduces in the Peer-to-Peer Network P2P flow medium live system is characterized in that, disposes medium source nodes in system, is used for the managing medium resource of whole system; And more than one cache routing server, the managing medium resource, the route that are used in the local field are calculated and the peer client-side management; And comprise:
After the stream medium data of receiving the peer client received request, the cache routing server was selected suitable peer client in other peer clients of managing, and the stream medium data between foundation and the request peer client is mutual;
When the media stream data of asking the peer client to need did not exist in the peer of this cache routing server admin client, the cache routing server was pushed to request peer client to the required stream medium data information of other cache routing server request;
When the cache routing server obtained required stream medium data failure from other cache routing server, the cache routing server obtained desired data from medium source nodes, is pushed to request peer client.
2. method according to claim 1, it is characterized in that, the method also comprises: medium source nodes sends the multidate information of self-administered cache routing server list to each cache routing server, and timed sending inquires that heartbeat message that whether this cache routing server survive is to corresponding cache routing server;
Cache routing server response self heartbeat message comprises bandwidth and the ability value coded message of self to medium source nodes in heartbeat message;
The cache routing server list that medium source nodes is safeguarded according to the information updating of responding.
3. method according to claim 2 is characterized in that, the method also comprises: the cache routing server upgrades in local peer client side list for the adding of the peer client of managing and the behavior of leaving;
The cache routing server is processed the peer client side list of managing and this cache routing server according to the behavior of peer client result sends to medium source nodes.
4. method according to claim 1 is characterized in that, fashionable when there being new peer client to add, the method further comprises: the principle that medium source nodes is preferential according to the region is assigned to corresponding cache routing server with new peer client;
New peer client adds to the cache routing server request, the cache routing server is according to the channel of this new peer client-requested, in local cache, adopt time coordinate to retrieve needed media resource, if there is current in progress media resource, the stream medium data fragment is pushed to new peer client;
If required media resource does not exist, then to medium source nodes request media resource, the cache routing server obtains the stream medium data frag info from medium source nodes by the mode of drawing;
After the cache routing server gets access to media resource, be pushed to the peer client of request, upgrade simultaneously the media resource buffer memory of safeguarding.
5. method according to claim 1 is characterized in that, the method also comprises: between each cache routing server the peer client side list of managing is undertaken by heartbeat message alternately.
6. method according to claim 1, it is characterized in that, described cache routing server is selected suitable peer client in other peer clients of managing, the stream medium data between foundation and the request peer client also comprises alternately: the peer client is calculated the ability value of the current disposal ability that shows this peer client according to current bandwidth and CPU usage; The peer client is attached to own ability value and BM block message and sends to the cache routing server in the heartbeat message;
When the stream medium data of receiving the peer client receives request, the cache routing server is according to the described ability value and the BM block message that obtain, adopt Hierarchical Matching Algorithm to carry out the suitable peer client of route calculative determination, and route result of calculation is sent to described request peer client.
7. method according to claim 6 is characterized in that, described route computational process comprises:
A. according to the order incremental of the client ip network segment of living in IP index of definition to every section network of database, for the multilayer network segment, successively set up index successively until top from minimum IP network section;
B. the index IP that sets up with client ip and steps A compares coupling, judges near distance, according to the division of existing IP four hop count values, from high section successively coupling until low section;
C. according to network layer match index in turn, from top successively deeply, until lowermost layer, the final selected zone of dividing, and this IP carried out in addition network area sign;
D. preferentially mate according to the area flag position;
E. with matching result ordering, according to ranking results, one by one the BM piece is compared, therefrom select to have the forward peer client of required BM piece and IP matching addresses as route results.
8. method according to claim 1, it is characterized in that, when the data that the peer client of described cache routing server admin needs do not exist in the peer of this cache routing server admin client, further comprise to the process of other cache routing server request data:
If the cache routing server load of requested date surpasses predefined threshold value, then this requested cache routing server is selected a peer client from the peer client of managing, and returns to request peer client as route results.
9. method according to claim 8, it is characterized in that, when described other cache routing server and the peer client of managing thereof do not have desired data or do not provide data, services, the cache routing device is also judged under the peer client: if the current media source node then provides data, services by medium source nodes under load threshold; If the current media source node surpasses load threshold, then denial of service.
10. method according to claim 1, it is characterized in that, described cache routing server comprises from the situation that other cache routing server obtains required stream medium data failure: other cache routing server and the peer client of managing thereof do not have desired data or data, services are not provided.
11. a realization system that reduces based on time-delay in the Peer-to-Peer Network P2P flow medium live system of each described method among the claim 1-10 is characterized in that, comprising:
Medium source nodes is for the managing medium resource of whole system;
More than one cache routing server, the managing medium resource, the route that are used in the local field are calculated and node administration; Described cache routing server is selected suitable peer client in other peer clients that it is managed, the stream medium data between foundation and the request peer client is mutual; When the media stream data of asking the peer client to need did not exist in the peer of this cache routing server admin client, the cache routing server was pushed to request peer client to the required stream medium data information of other cache routing server request; When the cache routing server obtained required stream medium data failure from other cache routing server, the cache routing server obtained desired data from medium source nodes, is pushed to request peer client.
12. system according to claim 11 is characterized in that, described medium source nodes further comprises:
The district management unit according to number of users and distributional region, creates and management domain, according to cache routing server in the selection of device resource characteristics and the specified domain, obtains the resource statistics information in each territory; Determine the merge and split in territory according to the quantity of member in the territory;
The node administration unit, by with OSS user authentication module interface, peer ID, the payment information of authentication peer client are guaranteed its legitimacy; According to the terminal characteristics, specify it to add the domain of dependence; Server, the peer resource of management the whole network;
Route computing unit carries out Preliminary screening to the whole network resource data, regularly to the overseas routing iinformation of cache routing server push; According to cache routing server route requests, inquire about the whole network resource statistics information of its storage, select to the relevant routing iinformation of other cache routing server request;
The network management interface unit has the platform management module interface in OAMAgent Agent and the OSS, realizes management, maintenance software upgrading.
13. system according to claim 11 is characterized in that, the cache routing server comprises:
The node resource administrative unit, the resource information of peer client, overseas resource information in storage and the management domain;
The node administration unit, withdraws from the adding of peer client in the management domain; By with the territory in the communication interface of peer client, regularly with the communication of peer client, monitor the peer client state;
Route computing unit with the medium source nodes communication, obtains up-to-date resource routing information; According to the feature of peer client, to the primary Calculation of resource data, regularly to peer client push routing iinformation; To take into account the territory internal burden during calculating balanced; The route requests of peer client represents the overseas route that the peer client please be looked for novelty to medium source nodes in response and the Acknowledge;
The network management interface unit has the platform management module interface in webmaster Agent and the OSS, realizes management, maintenance, the software upgrading of cache routing server.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101487269A CN101938508B (en) | 2009-07-01 | 2009-07-01 | Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101487269A CN101938508B (en) | 2009-07-01 | 2009-07-01 | Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101938508A CN101938508A (en) | 2011-01-05 |
| CN101938508B true CN101938508B (en) | 2013-01-02 |
Family
ID=43391636
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009101487269A Active CN101938508B (en) | 2009-07-01 | 2009-07-01 | Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101938508B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185880B (en) * | 2011-03-08 | 2014-06-18 | 天脉聚源(北京)传媒科技有限公司 | Video live-broadcast load equalizing method and video live-broadcast load equalizing system |
| CN102130838B (en) * | 2011-04-02 | 2013-09-25 | 南京邮电大学 | Method for realizing peer-to-peer network caching system based on Ares protocol |
| CN102510411A (en) * | 2011-12-28 | 2012-06-20 | 南京邮电大学 | Implementation method of cache server for Ares network |
| CN102857547B (en) * | 2012-04-12 | 2016-03-30 | 华为技术有限公司 | The method and apparatus of distributed caching |
| CN103067517B (en) * | 2012-12-31 | 2016-03-30 | 网宿科技股份有限公司 | In conjunction with being redirected and the P2P caching server system be initiatively connected and implementation method |
| CN103354545B (en) * | 2013-06-24 | 2015-04-15 | 西安交通大学 | Cloud computing based P2P streaming media server cluster deploying method |
| EP2819379A1 (en) | 2013-06-28 | 2014-12-31 | Thomson Licensing | Method for adapting the downloading behavior of a client terminal configured to receive multimedia content, and corresponding terminal |
| CN103945279B (en) * | 2014-05-17 | 2017-04-19 | 中国传媒大学 | Method for dynamic start time delay of P2P live broadcast streaming media system based on interruption density |
| CN105656978B (en) * | 2014-12-04 | 2019-06-25 | 中国移动通信集团公司 | A kind of resource share method and device |
| CN105162881B (en) * | 2015-09-25 | 2018-09-21 | 中铁工程装备集团有限公司 | A kind of P2P optimized transmission methods based on multimode terminal |
| CN107295037B (en) * | 2016-03-31 | 2021-06-01 | 中移(苏州)软件技术有限公司 | Live broadcast peer-to-peer network construction method and device based on software defined network |
| WO2017209573A1 (en) * | 2016-06-03 | 2017-12-07 | Samsung Electronics Co., Ltd. | Multi-point content transmission method and apparatus |
| CN107517229A (en) * | 2016-06-16 | 2017-12-26 | 阿里巴巴集团控股有限公司 | Generation, transmission method and the relevant apparatus of a kind of time source-routed information |
| CN106209860B (en) * | 2016-07-13 | 2021-07-30 | 杭州施强教育科技有限公司 | Real-time classroom streaming media live broadcast load distribution method |
| CN106254554A (en) * | 2016-09-30 | 2016-12-21 | 上海斐讯数据通信技术有限公司 | The file-sharing closed loop system realized based on Wifi storage and control method |
| CN107968755B (en) * | 2016-10-20 | 2021-04-02 | 上海交通大学 | Routing method and system for storing and forwarding media content based on MMT protocol |
| CN108173897B (en) * | 2016-12-07 | 2020-09-08 | 北京云中融信网络科技有限公司 | Request message processing method and device |
| CN107819704A (en) * | 2017-11-24 | 2018-03-20 | 量子云未来（北京）信息科技有限公司 | A kind of wireless extendible media application edge calculations device, system and method |
| CN108881963A (en) * | 2018-05-30 | 2018-11-23 | 歌尔科技有限公司 | Data capture method, server-side and client |
| CN112134955B (en) * | 2020-09-23 | 2024-03-22 | 北京达佳互联信息技术有限公司 | Data processing system, data processing method, data processing device, and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101068336A (en) * | 2006-08-22 | 2007-11-07 | 腾讯科技（深圳）有限公司 | Coordinate connection stream media direct broadcasting system and device |
| CN101127619A (en) * | 2007-09-29 | 2008-02-20 | 华中科技大学 | A P2P living broadcast method based on super node |
| WO2008043092A1 (en) * | 2006-10-05 | 2008-04-10 | Bittorrent, Inc. | Peer-to-peer streaming of non-live content |
-
2009
- 2009-07-01 CN CN2009101487269A patent/CN101938508B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101068336A (en) * | 2006-08-22 | 2007-11-07 | 腾讯科技（深圳）有限公司 | Coordinate connection stream media direct broadcasting system and device |
| WO2008043092A1 (en) * | 2006-10-05 | 2008-04-10 | Bittorrent, Inc. | Peer-to-peer streaming of non-live content |
| CN101127619A (en) * | 2007-09-29 | 2008-02-20 | 华中科技大学 | A P2P living broadcast method based on super node |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101938508A (en) | 2011-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101938508B (en) | Method and system for shortening time delay in peer-to-peer network streaming media live broadcast system | |
| US11539768B2 (en) | System and method of minimizing network bandwidth retrieved from an external network | |
| US20230164397A1 (en) | Distributed election-based end-to-end content distribution network system and distribution method | |
| US9497035B2 (en) | Method, device, and system for playing media based on P2P | |
| CN102355448B (en) | Cloud streaming media data transmission method and system | |
| KR101485610B1 (en) | Distributed content delivery system based on network awareness and method thereof | |
| Liu et al. | Locality-awareness in BitTorrent-like P2P applications | |
| WO2010105505A1 (en) | Method, user node and server for requesting position information of resource on network | |
| Trossen et al. | Towards an information centric network architecture for universal internet access | |
| US11843649B2 (en) | System and method of minimizing network bandwidth retrieved from an external network | |
| Shehab et al. | Efficient schemes for playout latency reduction in P2P-VoD systems | |
| Maly et al. | Comparison of centralized (client-server) and decentralized (peer-to-peer) networking | |
| Dai et al. | The disparity between P2P overlays and ISP underlays: issues, existing solutions, and challenges | |
| Wang et al. | Peer-assisted social media streaming with social reciprocity | |
| CN103179191A (en) | P2P (peer to peer) network control device and P2P network control architecture | |
| Johnsen et al. | Peer-to-peer networking with BitTorrent | |
| CN101369915B (en) | P2P operating network resource management system | |
| Alimi et al. | An open content delivery infrastructure using data lockers | |
| Al-Oqily et al. | Towards automating overlay network management | |
| Deltouzos et al. | SeekStream: adapting to dynamic user behavior in P2P video‐on‐demand | |
| Ellouze et al. | Collaboration between networks and applications in the future internet | |
| Zeng et al. | Enhanced video streaming network with hybrid P2P technology | |
| Li et al. | Dynamic transmission scheduling for streaming applications via P2P overlay | |
| Guo et al. | A novel P2P IPTV system for IPv4/v6 networks | |
| Abu-Dagga | BitTorrent Discovery and Performance Enhancement using DDS QoS Policies |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |