CN101047828B - Distribution conference system - Google Patents
Distribution conference system Download PDFInfo
- Publication number
- CN101047828B CN101047828B CN2006100670668A CN200610067066A CN101047828B CN 101047828 B CN101047828 B CN 101047828B CN 2006100670668 A CN2006100670668 A CN 2006100670668A CN 200610067066 A CN200610067066 A CN 200610067066A CN 101047828 B CN101047828 B CN 101047828B
- Authority
- CN
- China
- Prior art keywords
- node
- mcu
- mcu node
- conference system
- child
- 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
- Telephonic Communication Services (AREA)
Abstract
This invention provides a distribution conference system including at least a MCU node and its sub-nodes, in which, the MCU node is the voice-mixing node and signaling control node in the system, every MCU node mixes voices of the medium data sent by its sub-nodes and sends them to other MCU nodes directly or indirectly and sends the data with mixed voice from other MCU nodes to its sub-nodes, and the MCU node is played by user terminals, the selection and its number are determined according to numbers of attendees of a conference and its load ability, in which, each MCU node carries out load adjustment among MCU nodes according to the signaling information from other MCU nodes.
Description
Technical field
The present invention relates to the communications field, particularly a kind of distributed conference system.
Background technology
Along with popularizing of network, people no longer only are the exchanges of voice to the requirement of communication mode, but require to satisfy video and data synchronization transmission.And networks development also makes distributed office become universal phenomenon, and the people in the different location exchange by network, and this has promoted the demand to video conference service.The method of existing video conference mainly contains multicast and mixes dual mode.Though multicast does not require hand-in-glove between the participant, the advantage that the member can independently add or leave; But because network and router burden are heavy, poor stability is vulnerable to shortcomings such as attack, so existing a lot of networks are not supported multicast.The mode of mixing needs the server of central authorities and finishes work such as tissue and audio mixing, and is uploaded to all users, and this mode causes the ability of server and bandwidth to become bottleneck (Fig. 1).
For overcoming the above problems, the someone has proposed full distributed conference system (Fig. 2).In this system, each user connects and transceive data with other all users.The major defect of the method is: the signaling complexity, each user add and when withdrawing from all will with everyone exchange message; Requiring each user all to have enough bandwidth communicates by letter with all participants.Though Skype supports distributed multi-party communication in addition, what adopt is private technology.
Prior art document related to the present invention:
I.Jonathan?Lennox?and?Henning?Schulzrinne.“A?protocol?for?reliabledecentralized?conferencing，”NOS?SDAV?2003.
II.R.Boivie?et?al.“Explicit?multicast(xcast)basic?specification，”Internetdraft，IETF，Jan.2003.Work?in?progress.
III.O.Levin?and?P.K.Even.“High?level?requirements?for?tightly?coupled?SIPconferencing，”Internrnet?draft，IETF，Mar.2003.Work?in?progress.
Summary of the invention
In view of this, the object of the present invention is to provide the distributed conference system of a kind of network stabilization, load balancing.
According to distributed conference system of the present invention, the child node that comprises at least one MCU node and MCU node subordinate, the MCU node is audio mixing node and the signaling control node in this distributed conference system, the media data audio mixing that each MCU node sends its subordinate's child node also sends to other MCU node directly or indirectly with the data of audio mixing, and will send to its subordinate's child node from the audio mixing data of other MCU node, the MCU node is served as by user terminal, the selection of MCU node and MCU interstitial content are dynamically determined according to convention goer's the number and the load capacity of node, wherein, each MCU node is according to the load regulation of carrying out from the signaling message of other MCU node between the MCU node.
In aforesaid distributed conference system, establishing signaling between per two MCU nodes connects, or the signaling connection is connected with medium, when a MCU node and other all the MCU nodes in this distributed conference system have been set up medium and have been connected, this MCU node has been set up full connection, when a MCU node not with this distributed conference system in other all MCU nodes when setting up medium and being connected, the non-full connection of this MCU node.
In aforesaid distributed conference system, preserve network topology in each MCU node, and the node that writes down load capacity maximum in its child node is as the spare MCU node.
In aforesaid distributed conference system, when the network configuration of this distributed conference system changes, with the relevant MCU node updates network topology of network configuration change, concurrent SCN Space Cable Network topology updating message is given other MCU node in this distributed conference system, to upgrade the network topology in other MCU node.
In aforesaid distributed conference system, in network configuration owing to the adding of node or leave when changing or when arriving predetermined network configuration and regulating the time, the MCU node is carried out described load regulation.
In aforesaid distributed conference system, when network configuration changed owing to the adding of node, current MCU node determined that according to the load capacity of network present load and new node it is as new MCU node or as the child node of current MCU node.
In aforesaid distributed conference system, when new node adds this distributed conference system and during as new MCU node, current MCU node sends to this new MCU node with network topology, and present load is shifted this new MCU node near the MCU node of threshold value with its part parton node in this distributed conference system, as the child node of this new MCU node.
In aforesaid distributed conference system, when a certain node request during as the child node of a certain MCU node, whether this MCU node determines it as its child node according to its maximum load capability and maximum child node number,
When the node of this request can not be as the child node of this MCU node, with the child node of this node as other MCU node with minimum load, or increase by a MCU node also with the child node of this node as this new MCU node, maybe the node that will ask is as new MCU node.
In aforesaid distributed conference system, when network configuration changes owing to leaving of MCU node, the MCU node that leaves specifies the spare MCU node as new MCU node, and with the child node of its part or all of child node transfer as new MCU node.
In aforesaid distributed conference system, can not be all during when the child node of the MCU node that will leave as the child node of new MCU node, other MCU node of MCU node request that this will leave is accepted remaining child node.
In aforesaid distributed conference system, after the MCU node that will leave had shifted child node, it stopped and being connected of all MCU nodes according to predesigned order.
In aforesaid distributed conference system, the described network configuration adjusting time comprises that first network configuration is regulated the time and second network configuration is regulated the time, wherein, first network configuration adjusting time is set up the full back setting that connects by each MCU node at it, and second network configuration is regulated the time by there being to set up the full MCU node setting that connects.
In aforesaid distributed conference system, when reaching first adjusting during time, the MCU node determines whether to obtain child node as its child node from other MCU node according to the present load and the maximum load threshold value of this MCU node.
In aforesaid distributed conference system, present load at the MCU node does not surpass its maximum load threshold value, and its load is under the situation of the node of load minimum in all MCU nodes, and this MCU node obtains child node as its child node from other MCU node.
In aforesaid distributed conference system, when reaching for the first adjusting time, and the present load of MCU node surpasses under the situation of its maximum load threshold value, and this MCU node determines whether to adjust the network configuration of this distributed conference system according to the benefit of the cost of the ability of child node in this distributed conference system, regulating networks structure and regulating networks.
In aforesaid distributed conference system, when reaching second when regulating the time, current MCU node determines whether to exist and can set up the child node that is connected entirely with all MCU nodes, and when determining to have such child node, with this child node as the MCU node.
According to the present invention, have than the node of strength reason ability by the distributed earth Dynamic Selection and to finish focusing on of task as the MCU node, and each MCU node can be regulated its load according to the signaling message that transmits between the MCU node, therefore can make the load balancing of this conference system, and networking flexibility, and avoided the drawback of single failed point.
Description of drawings
Fig. 1 is the schematic diagram of the centralized conference system of prior art.
Fig. 2 is the schematic diagram of the full distributed conference system of prior art.
Fig. 3 is the schematic diagram according to distributed conference system of the present invention.
The flow chart of the processing that Fig. 4 carries out when being the initiation meeting.
The flow chart of the processing that Fig. 5 carries out when adding meeting for node.
The flow chart of the processing that current MCU node was carried out when Fig. 6 added meeting for node.
The flow chart of the processing that Fig. 7 carries out when leaving meeting for the MCU node.
The flow chart of the processing that other MCU node was carried out when Fig. 8 left meeting for the MCU node.
The flow chart of the processing that Fig. 9 carries out when receiving signaling message from other MCU node for the MCU node.
The flow chart of the processing that Figure 10 carries out in all cases for child node.
Figure 11 is the flow chart of the processing carried out when arriving the first adjusting time.
Figure 12 is the flow chart of the processing carried out when arriving the second adjusting time.
Embodiment
In according to distributed conference system of the present invention, unfixing central processing node.Disposal ability user strong and the bandwidth abundance is dynamically elected as " MCU (multipoint control unit) node " among all participants, for example has three MCU nodes in distributed conference system as shown in Figure 3.The groundwork of MCU node is audio mixing and packet distribution, and the number of MCU node is according to participant's number and has dynamically definite than the number of strong capable nodes.If the user increases, when certain MCU node burden surpassed thresholding, new MCU node is selected to come out to share its workload.If number of users reduces, some MCU node is changed back ordinary node, to keep load balancing.When if the MCU node leaves, in the existing subscriber, select new MCU node and take over its work.
In according to distributed conference system of the present invention, the MCU node be the audio mixing node also be the signaling control point.Maintain signaling between all MCU nodes and connect, but direct media flow transmission is not necessarily arranged.Each MCU node carries out audio mixing with the data flow of all child nodes of its subordinate to be handled, and the Media Stream of audio mixing is sent to other MCU node.When not having direct media flow transmission between two MCU nodes, the Media Stream that need transmit between these two MCU nodes is transmitted by other MCU node.The media data audio mixing that each MCU node sends its subordinate's child node also sends to other MCU node directly or indirectly with the data of audio mixing, and will send to its subordinate's child node from the audio mixing data of other MCU node.As can be seen from Figure 3, having the node of above signaling of two-way or two-way or data flow transmission is exactly the MCU node.In addition, when initiating meeting, the promoter of meeting is an original M CU node.
Store the network topology of this distributed conference system in each MCU node, the MCU node according to this network topology can determine to set up MCU node that medium are connected, not set up MCU node that medium connect with this MCU node, the data of audio mixing are by what path transmission etc.In addition, each MCU node is collected all child node ability information of its subordinate.Especially, the MCU node is stored ability is the strongest in all child nodes of its subordinate node as secondary node.In addition, when some new adding child node disposal ability is stronger, MCU can carry out role exchange with this initiate child node.In addition, in according to distributed conference system of the present invention, can carry out the distribution of load between the MCU node.
Below, with reference to accompanying drawing, the method for this distributed conference system of description architecture.
The flow chart of processing performed when initiating meeting has been described as shown in Figure 4.When the meeting promoter initiates meeting, at first in step S401, with meeting promoter's node as the MCU node.
Then, in step S402, ask other node corresponding to add meeting.
Then, in step S403, judge whether the node that adds has higher ability in step S402.When if the ability of the MCU node that the energy force rate of definite this initiate node is current is low ("No" among the step S403), with the child node of this initiate node as this MCU node.Have the ability higher ("Yes" among the step S403) if determine this initiate node, then in step S404, this node become new MCU node, and current MCU node becomes the child node of this new MCU node than current MCU node.Then, in step S405, structure is set regulates time T j.
In processing as shown in Figure 4,, therefore only exist a MCU node just can satisfy the requirement of system because the meeting networking is just set up.When the meeting person who participates in was more and more, initiate node can become new MCU node, and can not replace existing MCU node.This situation is described in detail with reference to flow chart as shown in Figure 5.
As shown in Figure 5, when a certain node request added meeting, it at first sent in step S501 and joins request to the node in the conference network.
Then, in step S502, judge whether this request is accepted.If when in step S502, being rejected this request, then directly finish, the requestor can not add meeting.When this joins request when being accepted in step S502, proceed to step S503, determine whether this initiate node is designated as the MCU node, promptly according to network present load and this initiate node load capacity determine that whether this node is as the MCU node.
When in step S503, determining that this initiate node is not designated as the MCU node, proceed to step S504, this node is set up signaling with current MCU node and is connected with medium, as this MCU node subordinate's a child node.
When in step S503, determining that this initiate node is designated as the MCU node, proceed to step S505, this initiate node is appointed as the MCU node, and obtains network topology from current MCU node.
Then, in step S506, according to the network topology that step S505 obtains, attempt with this network topology in all MCU nodes set up signaling and be connected with medium.When existence does not have to set up the MCU node that medium are connected with this initiate node, in step S507, ask this MCU node that the one child node is transferred under this initiate node, as an one child node.In step S508, ask to set up medium once more and be connected then with this MCU node.
Then in step S509, judge whether this initiate node has been set up medium with all MCU nodes and be connected.When determining to proceed to step S514 when all MCU nodes have been set up medium and are connected, structural adjustment time T j is set.When determining not have and all MCU nodes when connecting, in step S510, further judge whether current connection has reached the capacity threshold of oneself.When determining not reach the capacity threshold of oneself, return step S506, repeated execution of steps S506-510.When determining to have reached the capacity threshold of oneself, proceed to step S512, network configuration is set regulates regularly Ts, and proceed to step S513.In step S513, the connected MCU node of this initiate node request passes on the audio mixing medium to the MCU node that is connected with its medium.
Certainly, in processing as shown in Figure 5, when initiate node is designated as the MCU node, also can make it replace existing a certain MCU node in the network.Whether initiate MCU node replaces existing MCU node to determine according to the load and the disposition of whole network.
When existing MCU node receives the request OK response (represent this new node accept the request for conference that the MCU node sends) of the node of making a fresh start, add request for conference, or, carry out processing as shown in Figure 6 from the adding request for conference that child node is transmitted.
At first, in step S602, judge whether the load of its current MCU node and the number of subordinate's child node have not reached Th1 and Th2, Th1 is the maximum load threshold value of MCU node, Th2 is the maximum child node number threshold value of MCU node, Th1 also has nothing in common with each other for different its numerical value of MCU node with Th2, and Th2 is less than or equal to Th1.
When the number of determining its present load and subordinate's child node has not reached Th1 and Th2 ("Yes" among the step S602), proceed to step S610, accept this node as its child node, and in step S609, upgrade network topology and network topology is broadcast to other MCU node.When the number of determining its present load and subordinate's child node has reached Th1 and Th2 ("No" among the step S602), proceed to step S603, check whether there is the MCU node that satisfies these two threshold value Th1 and Th2 simultaneously in other all MCU nodes.
When other MCU node of these two threshold values is satisfied in definite existence simultaneously, in step S604, this request transferred to the MCU node that has minimum relative load in these MCU nodes.
When determining not have other MCU node that satisfies these two threshold values simultaneously,, judge whether to exist the MCU node that only satisfies Th1 by the judgement of step S605 and S606.If current MCU node satisfies Th1, then proceed to step S610, if current MCU node does not satisfy Th1, but other MCU node proceeds to step S604 when satisfying Th1.
When in determining all MCU nodes, not having load (among the step S606 not), proceed to step S607, specify spare MCU node or this initiate node as the MCU node less than the MCU node of its maximum load.Then, in step S608, current MCU node is set up signaling with this new MCU node and is connected (if be the spare MCU node before this new MCU node, originally just existed signaling to connect between current MCU node and the spare MCU node, this step not then), and with network topology notify this new MCU node.Then, in step S609, current MCU node updates network topology, and the network topology that will upgrade is notified other MCU node in this conference system.
When a node in the existing network leaves this meeting, determine that at first this node is MCU node or ordinary node (being child node).When this node that leaves was child node, the direct connected MCU node of this child node disconnected.When this node that leaves was MCU node (below, be called and leave the MCU node), it was handled as shown in Figure 7.Simultaneously, other MCU node (below, be called existing MCU node) also to handle accordingly, it is handled as shown in Figure 8.
As shown in Figure 7, at first, in step S701, specify the spare MCU node to take over the task that this leaves the MCU node.
In step S702, to leave the MCU node and send the audio mixing medium to the spare MCU node, the concurrent message that is sent from out is to other all MCU nodes.
In step S703, determine this leaves whether there is child node under the MCU node.When determining that this leaves under the MCU node without any child node, proceed to step 710, stop and being connected of all MCU nodes according to predesigned order.Otherwise, proceed to step S704.
In step S704, judge whether to reach the Th1 of spare MCU node.When determining not reached the Th1 of spare MCU node, a child node will leaving the MCU node in step S705 is transferred to the spare MCU node and is connected, and returns step S703.Thereby,, be transferred to singly with the spare MCU node and be connected leaving child node under the MCU node, until the Th1 that has reached the spare MCU node by the cycling of step S703-805.When determining to have reached the Th1 of spare MCU node, proceed to step and proceed to step S706, remaining child node is transferred to other MCU node that other has enough abilities.
Be similar to step S704, in step S707, judge whether to reach the Th1 of all MCU nodes.
When the Th1 that determines all MCU nodes has reached and has been left some child nodes ("Yes" of step S708), this leaves the termination of MCU node and is left being connected of child node in step S709, and proceeds to step 710.When determining that this leaves when not had child node under the MCU node, or carried out after the step S709, left the MCU node and in step 710, stop and being connected of all MCU nodes according to predesigned order.
When Fig. 8 describes a certain MCU of leaving node in detail and leaves meeting, the processing of existing MCU node.
At first, in step S802, existing MCU node can not shift any node and give and to be left the MCU node, and existing MCU node is that new MCU node is preserved current utilized bandwidth, is connected with new MCU node being used for.
Then, in step S803, determine whether to receive and leave the child node transfer request that the MCU node sends.If do not receive the child node transfer request, then in step S807, directly stop and leave being connected of MCU node.When in step S803, receiving the child node transfer request, proceed to step S804, this existing MCU node determines whether it has enough surplus resources to accept this child node.If resource is not enough, then this request of refusal in step S808 stops in step S807 then and leaves being connected of MCU node.
If existing MCU node determines that it has enough surplus resources and accepts this child node, then is this child node reserved bandwidth in step S805, in step S806, set up signaling then and be connected with medium and be connected, and upgrade network topology with new child node.Then, in step S807, stop and leave being connected of MCU node.
In this distributed conference system, the signaling message that the MCU node receives from other MCU node comprises that mainly topology is upgraded, child node shifts or receive request and set up the medium connection requests with other MCU node.Below, with reference to Fig. 9, the processing when describing MCU in detail and receiving the signaling message that other MCU node sends.
When receiving the signaling message of topology renewal in step S901, the topology of the topology that this MCU node at first receives this in step S902 and himself storage compares.When the topology of the topology of determining to receive and himself storage not simultaneously, in step S903, upgrade the topology of its storage.
When in step S904 when request MCU node receives its transfer of request or receive the request of child node, if disagree with ("No" among the step S905), then in step S908, send refuse information to this request MCU node; If agree ("Yes" among the step S905), then in step S906, give request MCU node with OK message (message is accepted in request), specify the child node and the MCU node of request to set up signaling and is connected with medium and is connected (situation of transfer child node), or be new child node reserved bandwidth and set up signaling with this new child node and be connected with medium be connected (situation of reception child node).Then, in step S907, upgrade topology, and the topology that will upgrade sends to all the MCU nodes in this distributed conference system.The description in front of the determining step of step S905 is described, and is not describing at this.
When the medium connection request that receives in step S910 from other MCU node, this MCU node determines in step S911 at first whether it has the ability to set up medium with the MCU node of this request and be connected.When this MCU node determines that it does not have the ability and the MCU node of this request to set up medium to be connected ("No" among the step S911), send the MCU node of refuse information to this request at step S914.When this MCU node is determined its have the ability when the MCU node of this request is set up medium and is connected ("Yes" among the step S911), setting up medium at step S912 with the MCU node of this request is connected, and upgrade topology in step S913, and the topology that will upgrade sends to all the MCU nodes in this distributed conference system.
In this distributed conference system, the performed processing of child node mainly contains three kinds of following situations: be appointed as the MCU node, because leaving of MCU node and and network disconnect and between the MCU node, shift.Below, with reference to Figure 10, describe the processing of child node under these three kinds of situations in detail.
When child node was appointed as the MCU node in step S1001, it at first obtained network topology from main MCU node in step S1002.Then in step S1003 with topology in all MCU nodes connect, and in step S1004, set up signaling with the child node of all requests as much as possible and be connected with medium.
When child node in step S1005 because leaving of connected MCU node and when disconnecting with network, it stops in step S1006 and being connected of this MCU node, and restarts the timer of next adding meeting in step S1007.Situation in step S1005 is meant: connected MCU node leaves, and this child node can not be transferred to the situation of other MCU node.
When child node shifted between the MCU node in step S1008, it was set up signaling with the target MCU node and is connected with medium in step S1009, and disconnected with the current MCU node that is connected.Step S1008 leaves at current MCU node, takes place under the situation such as network configuration adjusting.
In this distributed conference system, after each MCU node is set up full the connection, one first network adjustment time T i can be set all.In addition, do not have the full MCU node that connects, one second network adjustment time T s can be set.When these network adjustment times arrive, network configuration is regulated, shown in Figure 11 and 12.
As shown in figure 11, when the first network adjustment time T i, at first in step S1101, determine oneself whether also to have child node.If because the adjustment or the leaving of node of network make this MCU node not have any child node, this MCU node will be converted to a common child node.When determining not have the residue child node in step S1101, selection has the MCU node that minimum operating load is connected with medium from all MCU nodes in step S1113, becomes to be its child node.Then other MCU node of request stops being connected with medium with the signaling of this MCU node in step S1114, and upgrades network topology in step S1115.
When in step S1101, determine having the residue child node, proceed to step S1102, whether the relative load of determining current MCU node less than Th2, the maximum child node number that Th2 can connect for this MCU node.
When the relative load of in step S1102, determining current MCU node during less than Th2, proceed to step S1103, determine whether the relative load of this MCU node is the node of load minimum in all MCU nodes.When the node of load minimum in the not all MCU node of relative load of in step S1103, determining current MCU node, proceed to step S1106, restart the timing of Tj.When the relative load of determining current MCU node in step S1103 is the node of load minimum in all MCU nodes, proceed to step S1104, judge from the MCU node of maximum load to obtain whether the load of current MCU node becomes maximum after the child node.When in step S1104 be judged as "Yes" the time, proceed to step S1106, restart the timing of Tj.When in step S1104 be judged as "No" the time, proceed to step S1105, obtain a child node from this MCU node with maximum load, as the child node of current MCU node, then in step S1106, restart the timing of Tj.
When the relative load of determining current MCU node in step S1102 is not less than Th2, proceed to step S1107, determine this MCU node whether exist ability higher, can make the child node of load＜Th2.If there is such child node, then further in step S1108, determine whether to regulate benefit Bj greater than regulating cost Cj.When in step S1108, determining to regulate benefit Bj greater than adjusting cost Cj, current MCU node and this child node role exchange.When determining that regulating benefit Bj is not more than adjusting cost Cj, end process is waited for the arrival of next time regulating time T j.Related interstitial content when regulating cost and being MCU node and child node role exchange, propagation delay time, interim decline of user's received signal quality and signaling transport overhead etc.; The adjusting benefit is the raising of adjustment back user's received signal quality and network stabilization etc.
When in step S1107, determining that there is not such child node in this MCU node, proceed to step S1109, determine whether other MCU node exists such child node.When there is such child node in other MCU node, further in step S1110, judge the cost of transfer that shifts this child node and whether regulate cost less than regulating benefit, promptly judge whether to be necessary this child node is regulated.Related interstitial content when regulating cost and being MCU node and child node role exchange, propagation delay time, interim decline of user's received signal quality and signaling transport overhead etc.; The adjusting benefit is the raising of adjustment back user's received signal quality and network stabilization etc.Cost of transfer is that child node shifts caused signaling transport overhead between the MCU node, time delay, and signal quality descends degradation temporarily.The specific implementation of cost of transfer, adjusting cost and adjusting benefit can specifically define according to the characteristics of heterogeneous networks.When defining this child node of necessary adjusting among the step S1110, this child node is switched at the MCU of this child node node and current MCU node to part, then in step S1112 Tj is set.When determining that in step S1109 there is not such child node in other MCU node, when determining in step S1109 that perhaps other MCU node exists such child node still to there is no need to regulate this child node, execution in step S1112 is provided with Tj, and waits for the arrival of next time regulating time T j.
As shown in figure 12, when the second network adjustment time limit Ts arrives, determine in step 1201 at first whether the current MCU node that does not have to set up full connection exists a child node, it can be set up with other all MCU nodes and be connected entirely.When determining to have such child node, this child node and the current MCU node role exchange that is connected this child node, promptly this child node becomes new MCU node, and current MCU node becomes the child node of this new MCU node, and these all MCU nodes of new MCU node and other are connected full the connection.
When determining that there is not such child node in current MCU node in step 1201, whether definite other MCU node exists and can set up the child node that is connected entirely with other all MCU nodes in step S1202.When in step S1202, determining not have such child node, proceed to step S1204, reset the second network adjustment time limit Ts.When in step S1202, determining to have such child node, proceed to step S1203, ask the MCU node switching child node of this child node, the promptly current MCU node that does not have to set up full connection selects a child node and this child node that satisfies condition to exchange arbitrarily.Then, in step S1204, reset the second network adjustment time limit Ts.
According to the present invention, have node than strength reason ability as the MCU node by distributed Dynamic Selection, finish focusing on of task.Therefore, have load balancing, networking flexibility, and avoided the characteristics of single failed point.
Though the present invention expresses in concrete mode; but be not in order to qualification the present invention, any those skilled in the art, without departing from the spirit and scope of the present invention; can do various changes and retouching, so protection scope of the present invention is as the criterion when looking the patent claim person of defining.
Claims (16)
1. distributed conference system, it comprises at least one MCU node and MCU node subordinate's child node, the MCU node is audio mixing node and the signaling control node in this distributed conference system, the media data audio mixing that each MCU node sends its subordinate's child node also sends to other MCU node directly or indirectly with the data of audio mixing, and will send to its subordinate's child node from the audio mixing data of other MCU node, the MCU node is served as by user terminal, the selection of MCU node and MCU interstitial content are dynamically determined according to convention goer's the number and the load capacity of node
Wherein, each MCU node is according to the load regulation of carrying out from the signaling message of other MCU node between the MCU node.
2. distributed conference system as claimed in claim 1, wherein,
Establish signaling between per two MCU nodes and connect, or signaling connects and is connected with medium,
When a MCU node and other all the MCU nodes in this distributed conference system have been set up medium and have been connected, this MCU node has been set up full connection, when a MCU node not with this distributed conference system in other all MCU nodes when setting up medium and being connected, the non-full connection of this MCU node.
3. distributed conference system as claimed in claim 2, wherein,
Preserve network topology in each MCU node, and the node that writes down load capacity maximum in its child node is as the spare MCU node.
4. distributed conference system as claimed in claim 3, wherein,
When the network configuration of this distributed conference system changes, with the relevant MCU node updates network topology of network configuration change, concurrent SCN Space Cable Network topology updating message is given other MCU node in this distributed conference system, to upgrade the network topology in other MCU node.
5. as each described distributed conference system of claim 3-4, wherein,
In network configuration owing to the adding of node or leave when changing or when arriving predetermined network configuration and regulating the time, the MCU node is carried out described load regulation.
6. distributed conference system as claimed in claim 5, wherein,
When network configuration changed owing to the adding of node, current MCU node determined that according to the load capacity of network present load and new node new node is as new MCU node or as the child node of current MCU node.
7. distributed conference system as claimed in claim 6, wherein,
When new node adds this distributed conference system and during as new MCU node, current MCU node sends to this new MCU node with network topology, and present load is shifted this new MCU node near the MCU node of threshold value with its part parton node in this distributed conference system, as the child node of this new MCU node.
8. distributed conference system as claimed in claim 5, wherein,
When a certain node request during as the child node of a certain MCU node, whether the node that this MCU node is determined this request according to the maximum load capability and the maximum child node number of this MCU node as the child node of this MCU node,
When the node of this request can not be as the child node of this MCU node, with the node of this request child node as other MCU node with minimum load, or increase the child node of a MCU node and this new MCU node of the node that will ask conduct, maybe the node that will ask is as new MCU node.
9. distributed conference system as claimed in claim 5, wherein,
When network configuration changed owing to leaving of MCU node, the MCU node that leave specified the spare MCU node as new MCU node, and with the child node of its part or all of child node transfer as new MCU node.
10. distributed conference system as claimed in claim 9, wherein,
Can not be all during when the child node of the MCU node that will leave as the child node of new MCU node, other MCU node of MCU node request that this will leave is accepted remaining child node.
11. distributed conference system as claimed in claim 10, wherein,
After the MCU node that will leave had shifted child node, it stopped and being connected of all MCU nodes according to predesigned order.
12. distributed conference system as claimed in claim 5, wherein,
The described network configuration adjusting time comprises that first network configuration is regulated the time and second network configuration is regulated the time, wherein, first network configuration adjusting time is set up the full back setting that connects by each MCU node at it, and second network configuration is regulated the time by there being to set up the full MCU node setting that connects.
13. distributed conference system as claimed in claim 12, wherein,
When reaching first adjusting during time, the MCU node determines whether to obtain child node as its child node from other MCU node according to the present load and the maximum load threshold value of this MCU node.
14. distributed conference system as claimed in claim 13, wherein,
Do not surpass its maximum load threshold value in the present load of MCU node, and its load is under the situation of the node of load minimum in all MCU nodes, this MCU node obtains child node as its child node from other MCU node.
15. distributed conference system as claimed in claim 12, wherein,
When reaching for the first adjusting time, and the present load of MCU node surpasses under the situation of its maximum load threshold value, and this MCU node determines whether to adjust the network configuration of this distributed conference system according to the benefit of the cost of the ability of child node in this distributed conference system, regulating networks structure and regulating networks.
16. distributed conference system as claimed in claim 12, wherein,
When reaching second when regulating the time, current MCU node determines whether to exist and can set up the child node that is connected entirely with all MCU nodes, and when determining to have such child node, with this child node as the MCU node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2006100670668A CN101047828B (en) | 2006-03-31 | 2006-03-31 | Distribution conference system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2006100670668A CN101047828B (en) | 2006-03-31 | 2006-03-31 | Distribution conference system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101047828A CN101047828A (en) | 2007-10-03 |
| CN101047828B true CN101047828B (en) | 2011-05-25 |
Family
ID=38771987
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2006100670668A Active CN101047828B (en) | 2006-03-31 | 2006-03-31 | Distribution conference system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101047828B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101478619B (en) | 2009-01-05 | 2011-11-23 | 腾讯科技（深圳）有限公司 | Method, system and node device for implementing multi-path voice mixing |
| CN101515948B (en) * | 2009-03-31 | 2012-04-25 | 华为终端有限公司 | Management method, management device and management system in distributed multi-point conference system |
| CN101931782B (en) * | 2010-08-25 | 2015-08-12 | 中兴通讯股份有限公司 | For flow processing method and the device of multipoint control unit |
| CN102572368A (en) * | 2010-12-16 | 2012-07-11 | 中兴通讯股份有限公司 | Processing method and system of distributed video and multipoint control unit |
| CN103686059B (en) * | 2013-09-23 | 2017-04-05 | 广东威创视讯科技股份有限公司 | Distributed mixed audio processing method and system |
| WO2015117277A1 (en) * | 2014-02-10 | 2015-08-13 | 华为技术有限公司 | Method, apparatus and system for controlling multiparty real-time communication |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1297309A (en) * | 2000-02-01 | 2001-05-30 | 深圳市中兴通讯股份有限公司 | Method for increasing capacity of video conference system |
| WO2005004481A1 (en) * | 2003-07-07 | 2005-01-13 | Tandberg Telecom As | Distributed mcu |
-
2006
- 2006-03-31 CN CN2006100670668A patent/CN101047828B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1297309A (en) * | 2000-02-01 | 2001-05-30 | 深圳市中兴通讯股份有限公司 | Method for increasing capacity of video conference system |
| WO2005004481A1 (en) * | 2003-07-07 | 2005-01-13 | Tandberg Telecom As | Distributed mcu |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101047828A (en) | 2007-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101047531B (en) | Distributed conference system | |
| US7436428B2 (en) | System and method for including participants in a conference call | |
| CA2446707C (en) | Control unit for multipoint multimedia/audio system | |
| US6687234B1 (en) | Apparatus and method for optimal conference call renegotiation in telephony over LAN systems | |
| EP0805576A2 (en) | Multi-party communications | |
| CN101047828B (en) | Distribution conference system | |
| US8311197B2 (en) | Method and system for allocating, revoking and transferring resources in a conference system | |
| CN101453524B (en) | Multimedia service implementing method | |
| JPH0870321A (en) | Seamless multimedia conference system using improved multipoint control unit | |
| CN101656864A (en) | Method for realizing selection and watching of video and video session system | |
| EP2936803B1 (en) | Method and a device for optimizing large scaled video conferences | |
| JPH1198139A (en) | Conference management equipment, system and method for internet | |
| AU715048B2 (en) | A conference system | |
| CN102111387A (en) | Method for transferring control power in conference state and user equipment | |
| KR101026033B1 (en) | System and method for providing video service interworking multipoint conference unit | |
| CN113038064B (en) | Mobile terminal conference system | |
| US8068128B2 (en) | Visual communication server and communication system | |
| JP3748952B2 (en) | Multipoint connection network construction method and multipoint connection network system | |
| CN1960526B (en) | Method and system for implementing service of multiparty communication | |
| KR100622020B1 (en) | Method for controlling overlay multicast system | |
| KR100328553B1 (en) | Multicast method for a group's leader relaying | |
| US20070280221A1 (en) | Telephone system, telephone exchange apparatus, and route switching method | |
| EP2690825B1 (en) | Relay multicast transmission method and relay multicast transmission system | |
| CN114760488A (en) | Data transmission method, device, terminal equipment and storage medium | |
| CN101374067A (en) | Method for implementing son conference, network system and multimedia processor |
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 |