CN102075864A - MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure - Google Patents

MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure Download PDF

Info

Publication number
CN102075864A
CN102075864A CN2010106152067A CN201010615206A CN102075864A CN 102075864 A CN102075864 A CN 102075864A CN 2010106152067 A CN2010106152067 A CN 2010106152067A CN 201010615206 A CN201010615206 A CN 201010615206A CN 102075864 A CN102075864 A CN 102075864A
Authority
CN
China
Prior art keywords
node
nodes
message
path
multicast
Prior art date
Application number
CN2010106152067A
Other languages
Chinese (zh)
Inventor
周贤伟
张永德
张鸣鲁
彭莱
李永芳
林大泽
黄秋香
Original Assignee
青海西部矿业科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 青海西部矿业科技有限公司 filed Critical 青海西部矿业科技有限公司
Priority to CN2010106152067A priority Critical patent/CN102075864A/en
Publication of CN102075864A publication Critical patent/CN102075864A/en

Links

Abstract

The invention relates to a problem of minimizing the cost of a multicast routing forwarding structure in an AdHoc network, and construction of a delay limit multicast forwarding structure in a wireless network based on an MCDS (minimum connected dominating set) algorithm in a graph theory. According to particularity of a wireless signal transmission method, the cost and delay function in the wireless multicast routing are redefined. In the method for constructing the delay limit multicast forwarding structure based on the vertex coloring ideas in the graph theory, which is proposed based on the MCDS theory in the graph theory, by solving the MCDS, the purpose of constructing a multicast routing structure with the minimum cost is achieved, the concept of multicast routing delay limit is proposed, and the MCDS is constructed under the constraint. The algorithm is proved to be correct by the theoretical derivation, and the algorithm is proved to be effective by a lower approximate ratio compared with the similar algorithm. At the same time, the algorithm has a time complexity of O(n) and a message complexity of O(n), thus the high efficiency of the algorithm is proved further; and the algorithm has the advantage of adapting to flexible Adhoc networks.

Description

一种基于MCDS的延迟定界组播转发结构的构建方法 A booster structure-based method for constructing a delay multicast delimitation of MCDS

技术领域 FIELD

[0001] 本发明涉及无线网络通信技术领域,具体地说是涉及到路由选择对移动网络的动态适应和节省无线网络中无线广播发送的能量和带宽资源消耗问题的一种基于MCDS的延迟定界组播转发结构构建方法。 [0001] The present invention relates to wireless network communication technologies, and particularly to adapt to the dynamic routing of a mobile network and an energy saving of resources and bandwidth consumption issues a wireless network based on radio transmission delay is delimited MCDS multicast forwarding structure construction method.

[0002] [0002]

背景技术 Background technique

[0003] Ad Hoc网络组播转发结构分为树结构、mesh结构和混合结构3种类型。 [0003] Ad Hoc network multicast forwarding structure is divided into a tree structure, mesh structure and a mixed structure of three types. 在对组播树结构的研究中,普遍认为,组播的带宽消耗主要表现在节点对信息复制的份数上,即所建组播树的总边数上,要想最小化组播的带宽消耗,就要最小化组播树的边数,也即最小化组播树中引入的除源节点和目的节点之外的节点数,这就是网络中的Meiner树问题。 In a study of the multicast tree structure, generally agreed that the multicast bandwidth consumption mainly in the number of copies of the information node replication, on the side that is the total number of multicast tree built, in order to minimize bandwidth multicast consumption, it is necessary to minimize the number of sides of the multicast tree, i.e. the number of nodes other than the source node and the destination node to minimize the introduction of the multicast tree, which is the network Meiner tree problem. Pedro Μ. Ruiz等提出,在无线网络中,由于无线传输的广播特性,组播的带宽和能量消耗主要依赖于组播树中的转发次数,也即依赖于组播树中承担转发任务的节点个数。 Pedro Μ. Ruiz put forward, in a wireless network, due to the broadcast nature of radio transmissions, the multicast bandwidth and energy consumption mainly depends on the number of forwarding the multicast tree, i.e. in dependence forward packets to nodes in the multicast tree number. 因而,最小化ad hoc网络资源消耗的目标就是要建立一棵转发节点数最少的组播树,这个问题被归结为图论中寻找最小连通支配集(MCDS)问题。 Thus, the goal of minimizing the ad hoc network resource consumption is to establish a minimum number of nodes to forward multicast tree, the issue comes down to graph theory is to find the smallest connected dominating set (MCDS) problem. 另外,由于无线传输的广播特性,移动ad hoc网络路由技术需要高效可靠的网络广播技术支持,广播成了组播中常用的通信方式。 Further, since the broadcast nature of radio transmissions, the mobile ad hoc network routing technique requires highly efficient and reliable technical support broadcast network, broadcast multicast became commonly used in communication. 然而,有线网络中的简单泛洪广播不适用于ad hoc网络中带宽资源和节点能量稀缺的网络环境,因而冗余分组转发、邻近节点间的信道竞争和抢占、转发消息的碰撞和冲突等成为ad hoc网络需要解决的问题。 However, in the wired network is not available for simple flood broadcast an ad hoc network nodes and bandwidth resource scarce energy network environment, and thus redundant packet forwarding, and to seize the channel contention between adjacent nodes, the forwarding of messages and collision conflicts become ad hoc network problems to be solved.

[0004] 已有的路由协议可分为3种类型:先验式、被动式和混合式。 [0004] existing routing protocols can be classified into three types: a priori, passive and mixed. 这些类型的路由协议没有一种可以避免泛洪通信方式,而泛洪带来了一系列的广播风暴问题。 These types of routing protocols there a way to avoid flooding of communication, while the flooding has brought a series of broadcast storm problem. 一方面由广播风暴带来了高协议开销和对正在进行的通信造成干扰的后果;另一方面,泛洪是一种不可靠的通信方式,泛洪在被动式协议中会影响最短路径的选择,或是找不到任何本来应该存在的路径,在先验式协议中会造成全局信息陈旧从而影响路由更新的正确性。 On the one hand brings high protocol overhead and the consequences of interference with ongoing communications by the broadcast storm; on the other hand, flooding is not a reliable means of communication, flooding in passive agreement will affect the choice of the shortest path, or could not find any path that should exist in the prior agreements will result in the global information stale thus affecting the accuracy of routing updates. 近几年,提出了在Ad hoc网络中建立类似于有线网络中的骨干网络的虚拟骨干结构的方法。 In recent years, a method for establishing a virtual backbone structure similar to the wired network backbone network in Ad hoc networks. 运行在这种虚拟骨干结构上的先验式路由协议,在虚拟的网络中心节点中存储路由信息,用于更新、 修复路由、管理控制全局网络,部分被动式路由协议依靠虚拟骨干结构来代替路由发现中的全网泛洪机制,以此提高网络的可靠性和有效性。 Running on such a virtual backbone structure priori routing protocol, routing information stored in the virtual central node in the network, for updating, repairing routing, network management control of the situation, routing protocols rely on the passive portion of the backbone structure instead of a virtual route discovery the whole network flooding mechanism, in order to improve the reliability and efficiency of the network. 显然,减少虚拟骨干结构中的节点数目、找到一种在ad hoc网中有效建立虚拟骨干结构的方法是解决问题的关键。 Clearly, reducing the number of nodes of the virtual backbone structure, find an effective method for establishing a virtual backbone structure is the key to solve the problem in ad hoc networks. 目前提出的在Ad hoc网络中建立连通控制集(⑶S)的几种算法中,较为典型的是文献“Distributed Construction of Connected Dominating Set in Wireless Ad Hoc Networks,,中公开的一种分布式连通控制集的构建方法。文献中,将在ad hoc网络中建立虚拟骨干结构的问题等价为单元图中的连通控制集问题,从而在给定的节点集中寻找最小连通控制集(MCDS)。 Several algorithms based Ad hoc network communication control set (⑶S) currently proposed, it is typically a document "Distributed Construction of Connected Dominating Set in Wireless Ad Hoc Networks ,, disclosed in a distributed set of communication control construction methods in the literature, the establishment of an equivalent virtual backbone structure in the ad hoc network for communication with the control unit of FIG problems set in, so that a given set of nodes to find minimal connected dominating set (the MCDS).

[0005] Ad Hoc网络是一种没有固定网络基础设施的自组织网络。 [0005] Ad Hoc network is a self-organizing network is not a fixed network infrastructure. 它由一些移动节点组成,通过这些节点间的无线链路或中间节点的多跳链路相连保持网络连接和实现数据传递,因而具有分布式、自组织、多跳路由、动态拓扑、有限的安全保证等特点。 It consists of several mobile nodes via multi-hop link or a wireless link between the intermediate node connected to the nodes and maintain the network connection for data transfer, which has a distributed, self-organizing multi-hop routing, dynamic topology, limited security guarantee and so on. 近年来,研究人员对ad hoc网络组播已有了较深入的研究,按常用分类标准,ad hoc网络组播路由结构分为树结构、mesh结构和混合结构3种类型。 In recent years, researchers have been ad hoc network multicast a more in-depth research, according to commonly used classification criteria, the ad hoc network multicast routing structure is divided into a tree structure, mesh structure and hybrid structure 3 types. 在对组播树结构的研究中,普遍认为Steiner 树是最小代价组播树。 In a study of the multicast tree structure, Steiner tree is generally considered the minimum cost multicast tree. 但是,这一结论在无线组播通信中并不成立,Pedro Μ. Ruiz等指出,由于无线媒介的广播本质,(即只发送一个消息来转发组播报文给所有的下一跳节点, 而不是复制报文给下游邻居节点),组播通信的能量主要由节点转发消耗,总体取决于组播树中的报文转发次数,也即承担转发任务的节点个数。 However, this conclusion does not hold in the wireless multicast communication, Pedro Μ. Ruiz et al point out, due to the broadcast nature of the wireless medium, (that is, just send a message to forward multicast packets to all the next hop, rather than copy message to the downstream neighbor node), the energy mainly by multicast communication node forwards consumed, generally depending on the multicast tree packet forwarding number, i.e. the number of nodes assume the forwarding task. 因此,最小化ad hoc网络资源消耗的目标就是要建立一棵转发节点数最少的组播树,此问题归结为图论中寻找最小连通支配集(MCDS)问题。 Therefore, the goal of minimizing the ad hoc network resource consumption is to establish a minimum number of nodes to forward multicast tree, the problem boils down to find the minimum connected dominating set (MCDS) problem in graph theory.

[0006] [0006]

发明内容 SUMMARY

[0007] 本发明的目的是提供一种基于MCDS延迟定界组播转发结构的构建方法。 [0007] The object of the present invention is to provide a method for constructing forwarding delay Structure MCDS delimitation multicast.

[0008] 本发明一种基于MCDS延迟定界组播转发结构的构建方法通过下述技术方案予以实现:本发明一种基于MCDS的延迟定界组播转发结构的构建方法包括基于图着色的MIS选举和分布式组播树构建,所述的基于图着色的MIS选举为: [0008] The present invention is based MCDS delay method of constructing a multicast forwarding delimitation structure is realized by the following technical scheme: The present invention is based on constructing bounding method for delaying forwarding of multicast MCDS based MIS structure including graph coloring election and distributed multicast tree construction, the MIS-based electoral FIG coloring is:

源s知道到所有组播目的节点的最小时延路径 Source s to know the minimum delay path of all the multicast destination node

,找出所有目的节点中到源s的Pmirrtfi^中的最长路径目的节点,从该路径上与目的 , Longest path destination node to the destination node to identify the source of all the Pmirrtfi ^ s in, and the path from the object

节点相邻的节点开始,每相隔两跳确定一个独立集节点,最长路径如有k条,则执行k次直到找到所有独立集节点;这些独立集节点的确定过程中可能碰到如下两种情况: Start node adjacent to the node, determining a separate set of two-hop node, if k longest path section, each separated is performed k times until it finds all nodes independent set; determining the course of these separate sets of nodes may encounter the following two Happening:

1)后确定的独立集节点与之前确定的独立集节点相邻如图1所示,u为已确定的目的节点4的最小时延路径上的独立集节点,ν为另一目的节点6的最小时延等长路径上想要加入独立集的节点,因为L (P (s, 4» =L (P (s,《)),独立集节点之间的相隔为两跳,所以路径P(s,v)和路径H^w)长相差2的整数倍,即I聊,力)-L(P(_ \=2k,k = 0,1,2… ;如聊Λ)>2.,则节点u可以通过路径P(s,v)到达源节点s, LCPCCs, ^Cvj u)»L(P(js, v))+1<LCPCs, u)),与Ρ(β,ιί)为从u 到源s 的最短路径不符,u 与ν 互换,同理可知两路径长不可能相差两跳或两跳以上,所以。 1) independent set of nodes independent set of nodes is determined and the previously determined rear adjacent to 1, u is an independent set of nodes on the minimum delay path to the destination node 4 is determined, ν is a further object of the node 6 node minimum delay path length and the like want to add on an independent set, because the lag between L (P (s, 4 »= L (P (s,")), a two-hop node is set independently, the path P ( s, v) and the path H ^ w) differ by an integral multiple of length 2, i.e. I chat, force) -L (P (_ \ = 2k, k = 0,1,2 ...; as chat Λ)> 2,. the node u by a path P (s, v) reach the source node s, LCPCCs, ^ Cvj u) »L (P (js, v)) + 1 <LCPCs, u)), and Ρ (β, ιί) of does not match the shortest path from u to s of the source, u and ν interchangeably, similarly seen two possible path length difference between the two or more than two-hop hop so. LQKsM = L(JKsjl),将路径P(ν, 4 )通过节点u接入到P(s,P(v, 4))上,所以最大时延最多增加一; LQKsM = L (JKsjl), the path P (ν, 4) by an access node u to P (s, P (v, 4)), the delay is increased up to a maximum;

2)当P(s,dr )和P(s,4 )两路径在节点w处重叠时,由于I(P(Mr)) = I^Pferfi)), 2) When P (s, dr), and P (s, 4) overlap in the two paths at node w, since I (P (Mr)) = I ^ Pferfi)),

所以£(P(W,<)) = i(P(W,4)),两条路径在确定独立集节点时不会在节点w处发生冲突; Therefore £ (P (W, <)) = i (P (W, 4)), without conflict in the two paths at a node w in determining the set of independent nodes;

给这些独立集节点赋予最大优先级,其它节点按照权重大小赋予优先级,以分布式异步方式进行的着色过程将首先由这些已确定的MIS节点发起,节点基于邻居节点的信息确定自己的颜色:①最大优先级节点首先确定颜色②相邻节点的着色过程不能同时执行,优先级大的邻居节点具有优先着色权;初始化时,每个节点中有如下变量: color:节点颜色序号(初始化为0); P:优先级; Give these independent set of nodes given maximum priority, other nodes in the right major small given priority, coloring process in a distributed asynchronous manner will be first initiated by these MIS nodes have been identified, the node determines its own color based on information neighbor nodes: ① maximum priority node first determines the color of the coloring process ② neighbor nodes can not be performed simultaneously, priority neighbor node has a large priority right coloring; initialization, each node has the following variables: color: the color of the node number (initialized to 0 ); P: priority;

T :节点种类(M独立集节点;C连接节点)C:优先级高的邻居节点的颜色序号集合; T: the type of the node (M independent set of nodes; connecting node C) C: number of high priority color set of neighbor nodes;

MN. P:邻居MIS节点的优先级; . MN P: MIS priority neighbor node;

MN. N:邻居MIS节点的个数; . MN N: number of neighbor nodes of the MIS;

NF:优先级高的邻居节点的个数; NF: number of higher priority neighbor nodes;

S : {D,ND}节点状态(D已确定;ND未确定); S: {D, ND} node status (D have been identified; ND not determined);

节点收到邻居节点发来的D消息,执行如下算法: Node received from the neighbor node D sent message, perform the following algorithm:

1. If(S=ND) 1. If (S = ND)

2. { 2. {

3. if (D. Color=I) 3. if (D. Color = I)

4. { 4. {

5. MN. P=D. P ; . 5. MN P = D P.;

6. MN. N++; . 6. MN N ++;

7. if (MN. N > 2) 7. if (MN. N> 2)

T=C ; T = C;

8. } 8. }

9. else C D. color ; 9. else C D. color;

10. if (NF-=O) 10. if (NF- = O)

11. { 11. {

12. if (MN. N=O) 12. if (MN. N = O)

13. color=l; 13. color = l;

14. T=M ; 14. T = M;

15. else 15. else

16. for(i=2, color=0, i++) 16. for (i = 2, color = 0, i ++)

17. { 17. {

18. if (ie C) 18. if (ie C)

19. color=i; 19. color = i;

20. } 20}

21. 广播D消息给所有邻居节点(DP=P,D. color=color); 21. D broadcast message to all neighbor nodes (DP = P, D color = color.);

22. } twenty two. }

23. else 23. else

24. 发送probe消息给优先级大的邻居节点; 24. sends a probe message to neighbor nodes larger priority;

25. If(优先级大的节点的着色程序执行中) 25. If (large priority program execution node coloring)

26. 等待直到收到该邻居的D消息,跳转到程序开始执行; 26. Wait until the message is received D neighbor, jump to the beginning of the implementation of the program;

27. else 27. else

28. 跳转到程序的12行; 28. The program jumps to line 12;

29. }。 29.}.

[0009] 本发明一种基于基于MCDS的延迟定界组播转发结构的构建方法与现有技术相比较有如下有益效果:本发明提出一种基于MCDS构建延迟定界组播转发结构的方法,本发明基于网络的UDG模型,在实际的组播通信环境下,将MCDS看成组播转发结构,求解MCDS能减少组播通信代价。 [0009] The present invention method of constructing a structure-based forwarding delay delimitation based multicast MCDS compared with the prior art has the following advantages: provides a method for forwarding configuration MCDS based constructs of the present invention multicast delay delimitation, the present invention is based on the model of the network UDG, under actual environmental multicast communication, the multicast forwarding structure as MCDS solve MCDS to reduce the cost of multicast communication. 并有基于连通支配集的最小化时延广播策略的相关工作[12-13],还没有相关文献将求解MCDS的方法用于组播通信并考虑时延约束问题。 Communicating with a work based on the correlation and minimize latency broadcast policy dominating set of [12-13], there is no method for solving the literature MCDS for multicast communication and to consider the delay constraints. 本发明提出的基于图论中顺序点着色的时延定界组播路由算法,通过求解MCDS来实现构建最小代价组播路由结构的目的,并且在求解MIS阶段就考虑组播时延约束,基于组播目的节点的最小时延性能下限,对组播路由时延定界。 The present invention provides delay delimitation multicast routing algorithm in graph theory based Coloring order to achieve the object by solving the minimum cost MCDS constructing multicast routing structures, MIS and solving stage to take a multicast delay constraints, based on the lower limit of the minimum delay performance multicast destination node, delay delimit multicast routing.

附图说明 BRIEF DESCRIPTION

[0010] 本发明一种基于MCDS的延迟定界组播转发结构的构建方法有如下附图: 图1为该方法中Mis节点相邻的情况结构示意图; Construction Method [0010] Based on the delay delimitation MCDS multicast forwarding configuration of the present invention have the following drawings: where Mis a schematic view of the method of FIG. 1 for the adjacent nodes;

图2为该方法中MIS节点重叠的情况结构示意图; 图3为该方法中同一MIS节点发起的着色的情况结构示意图; 图4为该方法中不同MIS节点发起节点的着色的情况结构示意图。 2 a schematic configuration diagram of the case that process overlapping MIS nodes; FIG. 3 for a schematic view of the case of the method in the same node initiating MIS coloring; the structure of FIG 4 for a schematic coloring process initiating node MIS different nodes.

具体实施方式 Detailed ways

[0011] 下面结合附图和实施例对本发明一种基于MCDS的延迟定界组播转发结构的构建方法技术方案作进一步描述。 [0011] Next, the method of the present invention is constructed based on the delay MCDS multicast forwarding structure delimited aspect is further described in conjunction with the accompanying drawings and embodiments.

[0012] 如图1 一图4所示,本发明一种基于MCDS的延迟定界组播转发结构的构建方法包括如下步骤: [0012] As shown, the present invention is based on the delay delimitation MCDS multicast forwarding method of construction of a structure comprising the steps of a 4:

1. 基于图着色的MIS选举 1. MIS election based on graph coloring

源S知道到所有组播目的节点的最小时延路径^mM9 ,找出所有目的节点中到源S的 Source S to the minimum delay path of all known multicast destination nodes ^ mM9, find all destination node to the source S of

P—中的最长路径目的节点,从该路径上与目的节点相邻的节点开始,每相隔两跳确定 P- longest path in the destination node, the route starting from the destination node adjacent nodes, each separated by two-hop determination

一个独立集节点,最长路径如有k条,则执行k次直到找到所有独立集节点。 A set of independent nodes, if there are k longest path, then k times until it finds all nodes perform independent sets. 这些独立集节点的确定过程中可能碰到如下两种情况: The process of determining the set of independent nodes may encounter the following two cases:

1) 后确定的独立集节点与之前确定的独立集节点相邻如图1所示,U为已确定的目的节点4的最小时延路径上的独立集节点,V为另一目的节点·^的最小时延等长路径上想要加入独立集的节点,因为L (P (s— ))=L (PCs,《)),独立集节点之间的相隔为两跳,所以路径Pfev)和路径长相差2的整数倍,即I £<P(sM)-£<P(s,_h 2k,k=QX2... 。如ΚΡζε,φ- ΣΧΡ(β,ν))>2 ,则节点u可以通过路径Pfev)到达源节点s, L(P((s, ν),(ν, u));NL(P(s, v))+1<LCP(s, u)),与P(s,m)为从u 到源s 的最短路径不符,u 与ν 互换,同理可知两路径长不可能相差两跳或两跳以上,所以。 1) independent set of nodes independent set of nodes is determined and the previously determined rear adjacent 1, U is an independent set of nodes on the minimum delay path to the destination node 4 is determined, V is a further object of the nodes and ^ nodes want to join the long path independent set minimum delay, etc., as L (P (s-)) = L (PCs, ")), spaced between two independent sets of hop node, the path Pfev) and path length difference of an integral multiple of 2, i.e., I £ <P (sM) - £ <P (s, _h 2k, k = QX2 ... as ΚΡζε, φ- ΣΧΡ (β, ν).)> 2, the node u can be reached by a path Pfev) a source node s, L (P ((s, ν), (ν, u)); NL (P (s, v)) + 1 <LCP (s, u)), and P (s, m) of the shortest path from u to match the source s, u and ν interchangeably, similarly seen two possible path length difference between the two or more than two-hop hop so. 聊M))=聊为),将路径P(ν, 4 )通过节点u接入到P(s,P(v, 4 ))上,所以最大时延最多增加一。 Chat M)) = as chat), the path P (ν, 4) by an access node u to P (s, P (v, 4)), is increased up to a maximum delay.

[0013] 如图2所示,当P(s, 和P(s, 4 )两路径在节点w处重叠时,由于L(P(sA)) = L(P(s,dJ),所以Linw4;i) = L(PiwJt)),两条路径在确定独立集节点时不会在节点w处发生冲突。 [0013] As shown in FIG. 2, when the P (s, and P (s, 4) overlap in the two paths at node w, since L (P (sA)) = L (P (s, dJ), so Linw4 ; i) = L (PiwJt)), two paths without conflict in determining the set of independent nodes at node w.

[0014] 给这些独立集节点赋予最大优先级,其它节点按照权重大小赋予优先级,以分布式异步方式进行的着色过程将首先由这些已确定的MIS节点发起,节点基于邻居节点的信息确定自己的颜色:①最大优先级节点首先确定颜色②相邻节点的着色过程不能同时执行,优先级大的邻居节点具有优先着色权。 [0014] to these independent set of nodes given maximum priority, other nodes in the right major small given priority, coloring process in a distributed asynchronous manner will be first initiated by these MIS nodes have been identified, the node determines its own based on the information of neighbor nodes color: ① maximum priority node first determines the color of the coloring process ② adjacent nodes can not be performed at the same time, the priority of big neighbors have priority right coloring. 初始化时,每个节点中有如下变量: color:节点颜色序号(初始化为0); P:优先级; During initialization, each node has the following variables: color: the color of the node number (initialized to 0); P: priority;

T :节点种类(M独立集节点;C连接节点) T: the type of the node (M independent set of nodes; connecting node C)

C:优先级高的邻居节点的颜色序号集合; C: Color Reference high priority set of neighbor nodes;

MN. P:邻居MIS节点的优先级; . MN P: MIS priority neighbor node;

MN. N:邻居MIS节点的个数; . MN N: number of neighbor nodes of the MIS;

NF:优先级高的邻居节点的个数; NF: number of higher priority neighbor nodes;

S : {D,ND}节点状态(D已确定;ND未确定); S: {D, ND} node status (D have been identified; ND not determined);

节点收到邻居节点发来的D消息,执行如下算法: Node received from the neighbor node D sent message, perform the following algorithm:

1. If(S=ND) 1. If (S = ND)

2. { 2. {

3. if (D. Color=I) 3. if (D. Color = I)

4. { 4. {

5. MN. P=D. P ; . 5. MN P = D P.;

6. MN. N++; . 6. MN N ++;

7. if (MN. N a 2) 7. if (MN. N a 2)

T=C ; T = C;

8. } 8. }

9. else C D. color ; 9. else C D. color;

10. if (NF-=O) 10. if (NF- = O)

11. { 11. {

12. if (MN. N=O) 12. if (MN. N = O)

13. color=l; 13. color = l;

14. T=M ; 14. T = M;

15. else 15. else

16. for(i=2, color=0, i++) 16. for (i = 2, color = 0, i ++)

17. { 17. {

18. if (it C) 18. if (it C)

19. color=i; 19. color = i;

20. } 20}

21. 广播D消息给所有邻居节点(DP=P,D. color=color); 21. D broadcast message to all neighbor nodes (DP = P, D color = color.);

22. } twenty two. }

23. else 23. else

24. 发送probe消息给优先级大的邻居节点; 24. sends a probe message to neighbor nodes larger priority;

25. If(优先级大的节点的着色程序执行中) 25. If (large priority program execution node coloring)

26. 等待直到收到该邻居的D消息,跳转到程序开始执行; 26. Wait until the message is received D neighbor, jump to the beginning of the implementation of the program;

27. else28. 跳转到程序的12行; . 27. else28 program jumps to line 12;

29. } 29.}

2.分布式组播树构建 2. distributed multicast tree construction

目的节点按到源节点的最小时延路径的长短顺序依次寻路,接入组播树。 The destination node according to the order of the length of the minimum delay path to the source node sequentially pathfinding, to a multicast tree. 将k 条时延上界路径作为初始组播树T。 The upper bound on k pieces of delay path as an initial multicast tree T.

[0015] 步骤1树根s选择还未接入树T目的节中最长路径,该目的节点为将 [0015] Step 1 s roots has not selected the access object tree T longest path section, which is the destination node

要加入的节点,并通过该路径发消息S给它,激起该节点的寻路过程,该节点广播消息P。 Was added to the node, and the path through which the message S to it, stirred pathfinding process of the node, the node a broadcast message P. 当独立集节点转发消息P时,只有邻居节点中的连接节点(T=C)接收该消息参与转发,消息的来源节点不接收该消息。 When the independent set message forwarding node P, only the connection node (T = C) the neighbor node receives the participation message forwarding, message source node does not receive the message. 当连接节点转发该P消息时,只有邻居节点中的独立集节点(T=M) 接收该消息参与转发,消息的来源独立集节点或其它节点不接收该消息。 When the connection node P forwards the message, only the node neighbor nodes independent set (T = M) receives the participation message forwarding, independent sources of current nodes or other nodes do not receive the message. P消息中包括如 The message includes P

下信息:TTL= “―(该节点到源的最小时延路径长);PID :记录该消息经过路径上的节点ID。 The information: TTL = "- (the minimum delay of the source node to the path length); PID: recording the message on the path through the node ID.

[0016] 步骤2原树T上的MIS节点若收到消息P,将自己到树根的跳数加上PID中记录的节点个数,得到此条路径长的值L。 [0016] Step 2 MIS node on the tree T if the original message is received by P, and the number of hops to the root of their PID plus the number of nodes recorded in the entry path length obtained value L. 发送消息I,将L值和P消息中的PID信息递交给树根S。 Sending a message I, the L value and the PID information of the message submitted to the P roots S.

[0017]步骤 3 [0017] Step 3

a)树根s比较所有消息I中L和时延上界的大小,找出L值小于时延上界的I 消息,选出携带的PID信息记录的路径跳数最小的I消息,发送消息C,携带目的节点ID,首先按原树路径返回给选出的I消息的发送节点,然后根据PID中记录的节点经过这些节点到达要接入树T的目的节点,消息C经过的节点确定上一跳节点为父节点,下一跳节点为子节点,通过这种路径确认,将该路径接入到树T上。 Minimum a) s I, L Comparative roots and upper bounds of the size of all message latency to find the L value is less than the upper bound of the delay I message, select recording PID information carried hop-count I message, send the message C, ID of the destination node, the sending node first original tree path I returned to the selected message, then through the nodes to the access node to the destination node according to tree T is recorded in the PID, the node C passes the message to determine the hop node is a parent node, the child node to the next hop node, recognizing that this path, the path to the tree access T.

[0018] b)若树根s没有收到I消息或没有L值小于时延上界的I消息,通过目的节点的最小时延路径发送消息C,确认该条路径为目的节点的组播树接入路径。 [0018] b) if the root is not received I message or s is not less than the upper bound value L I delay message, send the message via C minimum delay path to the destination node, the acknowledgment in the multicast tree path is a destination node access path. 如果发生MIS节点冲突,按照MIS选举中步骤2中的方法将新路径接入到老路径上,最多增加一跳时延。 If MIS node collision occurs, according to the method of Step 2 election MIS new access path to the old path, increasing up hop latency.

[0019] 步骤4如有多个最小时延路径长相等的目的节点,按任意顺序重复执行以上步 [0019] Step 4 if a plurality of the minimum delay path looks like the destination node, in any order Repeat these steps

马聚ο Ma poly ο

[0020] 步骤5基于已建树T循环执行步骤1一4直到所有目的节点都接入树T。 [0020] Step 5 on T cycles step 4 has a contribution 1 until all nodes have access object tree T.

Claims (2)

1. 一种基于MCDS的延迟定界组播转发结构的构建方法,包括基于图着色的MIS选举和分布式组播树构建,其特征是所述的基于图着色的MIS选举为:源s知道到所有组播目的节点的最小时延路径iLw-,找出所有目的节点中到源S的iLwM5,中的最长路径目的节点,从该路径上与目的节点相邻的节点开始,每相隔两跳确定一个独立集节点,最长路径如有k条,则执行k次直到找到所有独立集节点;这些独立集节点的确定过程中可能碰到如下两种情况:1)后确定的独立集节点与之前确定的独立集节点相邻如图1所示,u为已确定的目的节点4的最小时延路径上的独立集节点,ν为另一目的节点力的最小时延等长路径上想要加入独立集的节点,因为L (P (s, 4)) =L (P (s,《)),独立集节点之间的相隔为两跳,所以路径Pfev)和路径P(s,w)长相差2的整数倍,即U(PfevJ)-£(P(s,2k,k = 0X2... ;如 1. A method of constructing a multicast forwarding delay delimited MCDS based structure, including MIS based electoral and distributed multicast tree graph coloring construct, wherein said MIS based electoral FIG coloring is: Source know s the minimum delay path of all the multicast destination nodes iLw-, find the longest path to the destination node to the destination node iLwM5 all source S, in, the destination node from the neighboring node on the path, every two separated determining a set of independent hop node, if k longest path item, is performed k times until it finds all nodes independent set; determining the course of these separate sets of nodes may encounter the following two cases: independent set of nodes determined) after 1 and independent set of nodes adjacent to the previously determined as shown in FIG, u is an independent set of nodes on the minimum delay path to the destination node 4 is determined, ν want to another object of the nodal force minimum delay paths of equal length 1 was added to separate sets of nodes, because the lag between the L (P (s, 4)) = L (P (s, ")), a two-hop node is set independently, the path Pfev) and the path P (s, w ) differ by an integral multiple of length 2, i.e., U (PfevJ) - £ (P (s, 2k, k = 0X2 ...; as Κβ(ε,ιφ_ ίΧβ(β,ν))>Ζ ,则节点u可以通过路径P(S3V)到达源节点s, LCP((s,灿,u))>=LCP(s,v))+l<L(P(s,u)),与Ρζβ-)为从u到源s的最短路径不符,u与ν互换,同理可知两路径长不可能相差两跳或两跳以上,所以;£(?汉_)=£(?叙力),将路径? Κβ (ε, ιφ_ ίΧβ (β, ν))> Ζ, the node u by a path P (S3V) reaching the source node s, LCP ((s, Chan, u))> = LCP (s, v)) + l <L (P (s, u)), and Ρζβ-) of the shortest path from u to match the source s, u and ν interchangeably, similarly seen two possible path length difference between the two or more than two-hop hop, so ; (? Han _) £ = £ (? Syrian forces), the path? (\ 4 )通过节点u接入到P(s,P(v, 4))上,所以最大时延最多增加一;2)当P(s,4)和卩(8,4 )两路径在节点w处重叠时,由于丛=所以= ,两条路径在确定独立集节点时不会在节点w处发生冲突;给这些独立集节点赋予最大优先级,其它节点按照权重大小赋予优先级,以分布式异步方式进行的着色过程将首先由这些已确定的MIS节点发起,节点基于邻居节点的信息确定自己的颜色:①最大优先级节点首先确定颜色②相邻节点的着色过程不能同时执行,优先级大的邻居节点具有优先着色权;初始化时,每个节点中有如下变量: color:节点颜色序号(初始化为0); P:优先级;T :节点种类(M独立集节点;C连接节点)C:优先级高的邻居节点的颜色序号集合;MN. P:邻居MIS节点的优先级;MN. N:邻居MIS节点的个数;NF:优先级高的邻居节点的个数;S : {D,ND}节点状态(D已确定;ND (\ 4) access to P (s, P (v, 4)), the maximum delay increases so that a maximum by a node u; 2) when P (s, 4), and Jie (8,4) two paths overlapping at node w, so = = since plexus, two paths without conflict in determining the set of independent nodes in the node w; independent set of nodes to which maximum priority given to other nodes in the right small priority is given significant to coloring process of distributed asynchronously first MIS initiated by these nodes have been identified, the node determines its own color based on information neighbor nodes: ① maximum priority node first determines the color of the coloring process ② adjacent nodes can not be performed at the same time, priority magnitude neighboring node having the right of priority coloring; initialization, each node has the following variables: color: the color of the node number (initialized to 0); P: priority; T: the type of node (M independent set of nodes; connecting node C of NF; number of neighbor nodes MIS::) C: color number of high priority set of neighbor nodes;:; MN P. MIS priority neighbor node MN N. the number of higher priority neighbor nodes; S: {D, ND} node status (D have been identified; ND 确定);节点收到邻居节点发来的D消息,执行如下算法: Determination); neighbor node receives the node message sent by D, perform the following algorithm:
Figure CN102075864AC00021
if (MN. N > 2) T=C ;}else C^-D. color ;if (NF-=O){if (MN. N=O) color=l; T=M ;elsefor (i=2, color=0, i++){if (ι - C) coior=i;}广播D消息给所有邻居节点(DP=P,D. color=color);}else发送probe消息给优先级大的邻居节点; If (优先级大的节点的着色程序执行中)等待直到收到该邻居的D消息,跳转到程序开始执行; else跳转到程序的12行;.29. } 。 if (MN N> 2.) T = C;} else C ^ -D color;. if (NF- = O) {if (MN N = O.) color = l; T = M; elsefor (i = 2 , color = 0, i ++) {if (ι - C) coior = i;} broadcast D message to all neighbor nodes (DP = P, D color = color);} else sends a probe message to the priority of a large neighbor nodes. ; the If (large priority program execution shader node) wait until a message is received the neighbor D, the program jumps to begin execution;} .29; else jump to the 12 lines of the program.
2.根据权利要求1所述的基于MCDS的延迟定界组播转发结构的构建方法,其特征是所述的分布式组播树构建为:目的节点按到源节点的最小时延路径的长短顺序依次寻路,接入组播树;将k条时延上界路径作为初始组播树T ;1)步骤1树根s选择还未接入树T目的节点的中最长路径,该目的节点为将要加入的节点,并通过该路径发消息S给它,激起该节点的寻路过程,该节点广播消息P ;当独立集节点转发消息P时,只有邻居节点中的连接节点T=C接收该消息参与转发,消息的来源节点不接收该消息;当连接节点转发该P消息时,只有邻居节点中的独立集节点T=M接收该消息参与转发,消息的来源独立集节点或其它节点不接收该消息;P消息中包括如下信息:TTL= “―该节点到源的最小时延路径长;PID :记录该消息经过路径上的节点ID ;2)步骤2原树T上的MIS节点若收到消 2. Construction of structure-based method of forwarding a multicast MCDS delay the delimitation of claim 1, wherein said distributed multicast tree constructed: the length of the destination node according to the minimum delay path source node pathfinding in sequence, to a multicast tree; the k pieces of delay path as the upper bound of the initial multicast tree T; 1) step 1 s roots has not selected an access destination node tree T longest path, the object node is the node to be added, and a message path S through which it aroused pathfinding process of the node, the node a broadcast message P; when the node forwards the message independent set P, the node connected only neighbor nodes T = C receives the message involved in forwarding source node messages does not receive the message; when a connection node forwards the P message, only the independent set node T neighbor node = M receives the message involved in forwarding sources independent set of nodes or other node does not receive the message; P message includes the following information: TTL = "- the minimum delay path length to the source node; PID: recording the message on the path through the node ID; 2) step 2 on the MIS original tree T If the node receives elimination 息P,将自己到树根的跳数加上PID中记录的节点个数,得到此条路径长的值L ;发送消息I,将L值和P消息中的PID信息递交给树根s ;3)步骤3a)树根s比较所有消息I中L和时延上界的大小,找出L值小于时延上界的I消息,选出携带的PID信息记录的路径跳数最小的I消息,发送消息C,携带目的节点ID,首先按原树路径返回给选出的I消息的发送节点,然后根据PID中记录的节点经过这些节点到达要接入树T的目的节点,消息C经过的节点确定上一跳节点为父节点,下一跳节点为子节点,通过这种路径确认,将该路径接入到树T上;b)若树根s没有收到I消息或没有L值小于时延上界的I消息,通过目的节点的最小时延路径发送消息C,确认该条路径为目的节点的组播树接入路径;如果发生MIS节点冲突,按照MIS选举中步骤2中的方法将新路径接入到老路径上,最多增加一跳时延;4) P information, the number of hops to the root of their PID plus the number of nodes recorded in this entry to obtain a long path value L; send message I, the L value and the PID information submitted message to the P roots S; minimum 3) step 3a) Comparative I, L s roots and upper bounds of the size of all message latency, latency to find the L value is less than the upper bound I message, select recording PID information carried in the message hop-count I , C sends a message, ID of the destination node, the sending node first original is returned to the tree path I selected message, then the node according to the PID recorded through these nodes to reach the destination node tree T of the access, message passing C hop node node determines the parent node, child node of a next hop node, the path through this confirmation, the access path to the tree T; b) if the root s message is not received or not I is less than the value L I bound on the delay message, send the message via C minimum delay path to the destination node, the multicast tree path acknowledgment in the access path to the destination node; MIS node if a collision occurs, according to the election step MIS method 2 the new access path to the old path, increasing up to jump delay; 4) 步骤4如有多个最小时延路径长相等的目的节点,按任意顺序重复执行以上步骤;5)步骤5基于已建树T循环执行步骤1一4直到所有目的节点都接入树T。 Step 4 If the minimum delay path looks like plurality of destination nodes, according to the above steps repeatedly performed in any order; 5) based on T cycles step 5 step 4 has a contribution 1 until all nodes have access object tree T.
CN2010106152067A 2010-12-30 2010-12-30 MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure CN102075864A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010106152067A CN102075864A (en) 2010-12-30 2010-12-30 MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010106152067A CN102075864A (en) 2010-12-30 2010-12-30 MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure

Publications (1)

Publication Number Publication Date
CN102075864A true CN102075864A (en) 2011-05-25

Family

ID=44034178

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010106152067A CN102075864A (en) 2010-12-30 2010-12-30 MCDS (minimum connected dominating set)-based method for constructing delay limit multicast forwarding structure

Country Status (1)

Country Link
CN (1) CN102075864A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103067285A (en) * 2012-12-25 2013-04-24 北京银易通网络科技有限公司 Energy-saving and reliable multicast method based on relay clique in mobile ad-hoc network and sensor network
CN104010342A (en) * 2014-05-07 2014-08-27 南京邮电大学 Wireless sensor network topology self-healing algorithm based on node neighbor relationships
RU2621604C1 (en) * 2016-02-02 2017-06-06 Федеральное государственное бюджетное образовательное учреждение высшего образования "Вятский государственный университет" Control method of mobile self-organizing network topology

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1917460A (en) * 2006-09-07 2007-02-21 上海交通大学 Distribution type structuring method for virtual backbone network in network of wireless sensor
US20080219158A1 (en) * 2007-03-09 2008-09-11 Nbc Universal, Inc. Media content distribution system and method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1917460A (en) * 2006-09-07 2007-02-21 上海交通大学 Distribution type structuring method for virtual backbone network in network of wireless sensor
US20080219158A1 (en) * 2007-03-09 2008-09-11 Nbc Universal, Inc. Media content distribution system and method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
彭莱等: "《ad hoc网络中基于MCDS构建延迟定界组播转发结构》", 《计算机应用研究》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103067285A (en) * 2012-12-25 2013-04-24 北京银易通网络科技有限公司 Energy-saving and reliable multicast method based on relay clique in mobile ad-hoc network and sensor network
CN104010342A (en) * 2014-05-07 2014-08-27 南京邮电大学 Wireless sensor network topology self-healing algorithm based on node neighbor relationships
CN104010342B (en) * 2014-05-07 2018-01-02 南京邮电大学 Radio sensing network topology Self-healing Algorithm based on nodes neighbors relation
RU2621604C1 (en) * 2016-02-02 2017-06-06 Федеральное государственное бюджетное образовательное учреждение высшего образования "Вятский государственный университет" Control method of mobile self-organizing network topology

Similar Documents

Publication Publication Date Title
Raman et al. Design and evaluation of a new MAC protocol for long-distance 802.11 mesh networks
Yuan et al. An optimized ad-hoc on-demand multipath distance vector (AOMDV) routing protocol
Butenko et al. On the construction of virtual backbone for ad hoc wireless network
Cardei et al. Connected Domination in Multihop Ad Hoc Wireless Networks.
CN101138207B (en) Hybrid mesh routing selection method and access point
Maric et al. Cooperative multicast for maximum network lifetime
Wightman et al. A3: A topology construction algorithm for wireless sensor networks
Acharya et al. A label-switching packet forwarding architecture for multi-hop wireless LANs
Tyagi et al. Performance analysis of proactive and reactive routing protocols for ad hoc networks
Yu et al. GROUP: A Grid-Clustering Routing Protocol for Wireless Sensor Networks.
CN102404817B (en) Networking method and system for wireless sensor network based on multiple sink nodes
CN102026329B (en) Wireless communication network and self-adaptive route selecting communication method thereof
CN101208909B (en) Transmission power control over wireless ad-hoc network
KR100617715B1 (en) Method for transmitting Flooding Ad hoc Traffic Indication Message in MANET and medium access control protocol layer module therefor
Shaikh et al. New metrics for dominating set based energy efficient activity scheduling in ad hoc networks
CN101197748A (en) Multi-jump wireless self-organizing network construction method based on partition tree
CN101005422A (en) Method for establishing radio sensor network rout ebased on route neighbour list
CN1731760A (en) Ad Hoc network routing method based on flooding mechanism
CN101217381A (en) Wireless transducer network energy saving method based on cross layers
Halford et al. Barrage relay networks
Kumar et al. A survey on scheduling algorithms for wireless sensor networks
CN101355517A (en) Method for balancing network load based on wireless sensor energy information
EP1598994B1 (en) Generating and updating a hierarchical tree structure in an ad-hoc communication network with a MANET multicast routing protocol.
Schaefer et al. Potentials of opportunistic routing in energy-constrained wireless sensor networks
EP1705850B1 (en) System and method for collecting routing information in a network based on a mesh topology

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)