CN102244840B - Method for routing multicasts and allocating frequency spectrums in cognitive wireless Mesh network - Google Patents

Abstract

The invention discloses a method for routing multicasts and allocating frequency spectrums in a cognitive wireless Mesh network, comprising the following steps of: modelling the cognitive wireless Mesh network composed of a static CR-Mesh router into an undirected multi-graph; then, using a genetic algorithm to obtain a channel allocation scheme that a maximum value of path delay is minimum, including initialization; based on a shortest path algorithm, calculating an individual fitness function value; executing selection, crossover and mutation operations; generating the next generation of population; achieving a generation number; and outputting a plurality of steps of an optimal solution. By means of the invention, the problem of routing multicasts and allocating frequency spectrums of a monophyletic node in the cognitive wireless Mesh network can be solved; therefore, the maximum value of the path delay from a source node to all multicast destination nodes is minimum; and the network performance can be effectively improved.

Description

Multicast path in a kind of cognitive radio Mesh network is by reaching frequency spectrum distributing method

Technical field

The invention belongs to the radio network technique field, relate in a kind of cognitive radio Mesh network multicast path by and frequency spectrum distributing method.

Background technology

(Wireless Mesh Network is a kind of static multi-hop transmission network WMN) to wireless Mesh netword, comprises the node of three types at Mesh gateway, Mesh router and Mesh terminal.Mesh router and Mesh gateway have constituted the backbone network of wireless Mesh netword.The Mesh router has routing function, and it has a plurality of wireless transceivers usually.The Mesh gateway not only has the function of Mesh router, and has with Internet by the function that wire cable directly links to each other, and is the bridge between wireless Mesh netword and the Internet.Mesh terminal use's data arrive the Mesh gateway by the multi-hop transmission of Mesh router.Equally, in down link, the data of Mesh gateway arrive the distribution of Mesh terminal realization data by the multi-hop transmission of Mesh router.Wherein multicast is exactly a kind of typical data transmission, and multicast also is that wireless Mesh netword is towards a kind of critical network business of practical application.Such as, video conference, long-distance education, content distribution etc. are that the typical case of Streaming Media multicast uses.

Frequency spectrum (channel) is non-renewable resources a kind of preciousness, limited, growing along with being on the increase of the continuous expansion of coverage, wireless user, access service, and assignable frequency spectrum resource is fewer and feweri.(Cognitive Radio, CR) technology is a kind ofly can effectively alleviate the deficient network of problem technology of frequency spectrum to cognitive radio.

Wireless Mesh netword is as broadband access system of future generation.Cognitive radio technology is applied to the problem that its frequency spectrum of solution lacks in the wireless Mesh netword has potential advantage.Cognitive radio Mesh network (Cognitive Wireless Mesh Network, CWMN) in, each Mesh node (comprising Mesh router, Mesh gateway, Mesh terminal) uses the CR technology.For a Mesh node (CR-Mesh router, CR-Mesh gateway, CR-Mesh terminal general designation CR-Mesh node) that is equipped with CR, it can the perception main system in untapped frequency spectrum, and dynamically be linked into these available spectrum.The present invention does not consider concrete frequency spectrum perception algorithm, be primarily aimed at the CR-Mesh node obtained under available channel and the relevant information condition multicast path by and the spectrum allocation may algorithm.

Multicast is important research field in the wireless Mesh netword as a kind of communication service that can effectively save Internet resources always.In recent years, along with the progress of bottom wireless communication technology and the further investigation of WMNs self, WMNs presents with many radio frequencies, multichannel and many speed (multi-radio multi-channel multi-rate, be called for short MR2-MC) be the brand-new network form of essential characteristic, and obtained some achievements in research.Channel allocation also is the research emphasis of wireless Mesh netword always, and has been proved to be NP completely, has also obtained some achievements in research.

But, achievement in research at multicast and channel allocation in the wireless Mesh netword can not directly apply in the cognitive radio Mesh network, multicast in the cognitive radio Mesh network and channel allocation have following main feature: 1) multicast protocol in the wireless Mesh netword is to be operated in the use fixed channel, and the communications status of channel is known, cognitive radio Mesh network is operated in the environment of dynamic change, 2) the CR-Mesh node uses channel must guarantee PU not to be produced to disturb, 3) the CR-Mesh node can with channel be a subclass of all channels in the wireless environment, and this subclass is dynamic change, 4) isomerism of the available channel in the cognitive radio Mesh network, namely the probability of use of these available channels is different generally speaking.

Multicast path in the cognitive radio Mesh network by with channel allocation be key issue to be solved.

Summary of the invention

Technical problem to be solved by this invention provide in a kind of cognitive radio Mesh network multicast path by and frequency spectrum distributing method, lower multicast by effective spectrum allocation may and postpone, improve network performance, improve the availability of frequency spectrum.

The technical solution of invention is as follows:

Multicast path in a kind of cognitive radio Mesh network is by reaching frequency spectrum distributing method, at first the cognitive radio Mesh network that static CR-Mesh router is formed is modeled as a undirected multigraph G=(V, E, D), wherein V represents the set of CR-Mesh router, E represents to link the set of the Radio Link of two CR-Mesh routers that can intercom mutually, and D represents the set of the transmission delay of Radio Link between two CR-Mesh router nodes, d _iThe transmission delay of representing a certain channel i, each node v _i∈ V has the set of available channels K of a perception _i, d _IjExpression node v _iWith node v _jBetween physical distance, all nodes adopt half-duplex mode work;

K={1,2 ... k}={1,2,3,4,5,6} represent the set of total available channel, Ψ _{I, j}Expression node v _iWith node v _jIdentical set of available channels, S represents multicast source node, R={r ₁, r ₂... r _mThe set of expression multicast destination node; G '=(V ', E ', D ') represent the simple and connected graph by the G derivation, wherein, V '=V,

Any two nodes only have a usable wireless link among the G ';

T=(V _T, E _T) the expression multicast tree, P _T(S, r _i) among the expression multicast tree T one from S to a certain destination node r _iThe path, d _LDelay among the expression multicast tree T on the Radio Link L, Delay (P _T(S, r _i)) among the expression multicast tree T from source node S to destination node r _iPath delay, namely $\mathrm{Delay}\left(P_T(S,r_i)\right)={\mathrm{\Σ}}_{L\∈P_T(S,r_i)}{d}_L;$

Take genetic algorithm to ask for the channel assignment scheme of the maximum minimum in path delay then, GEN represents genetic algebra, and step is as follows:

1) initialization genetic algebra counter g=0;

2) initialization population, at first set chromosomal coded representation mode, use length to show a chromosome as the 7 system string lists of n, representing a channel assignment scheme, n=|E ' wherein |=23, the limit number of the simple and connected graph G ' that E ' presentation graphs G derives is numbered from 1 to n to n bar limit among the multicast tree T, and coded representation is X=x ₁x ₂X _n, x _i∈ 0}U K, i ∈ 1,2 ..., n} is in chromosome, if x _i=0, any channel is not distributed on the limit that is numbered i in the expression multicast tree, and the delay on the Radio Link that this limit is represented will be infinity, if x _i=k, k ∈ K illustrates that then the channel of the limit distribution that is numbered i is k;

Chromosome in the population adds up to N, and chromosome j is expressed as C in the population _j=c ₁(j) c ₂(j) ... c _n(j), c _i(j) ∈ 0}U K, i ∈ 1,2 ..., n}, j ∈ 1,2 ..., N}, the initialization population comprises following steps:

I) initialization chromosome counting device j=0;

Ii) chromosome C _jBe initialized as sky, j ∈ 1,2 ..., N};

Iii) initialization chromosome C _jLimit counter i=0;

Iv) initialization chromosome C _jC _i(j) channel on limit is set chromosome C _jTwo nodes connecting of i bar limit be v _aAnd v _b, initialization c then _i(j) for belonging to Ψ _{A, b}If any value in the set is Ψ _{A, b}=Φ, then c _i(j)=0, i.e. hinged node v _aAnd v _bI bar limit do not distribute any channel;

V) the limit counter adds 1, the channel that next bar limit of expression initialization is distributed;

Vi) judge the size of i and n, determine whether the limit that initialization is all, if change step vii; Otherwise, change step I v;

Vii) the chromosome counting device adds 1, the next chromosome of expression initialization;

Viii) judge the size of j and N, determine whether the chromosome that initialization is all, if finish; Otherwise, change step I i;

3) calculating ideal adaptation degree value based on shortest path first is the fitness function value:

Find the solution source node S to the multicast tree of destination node set R based on shortest path first, the delay of Radio Link is designated as the weights on limit, ask the shortest path of source node S all destination nodes in the destination node set R, the peaked inverse in source node S path delay of all destination nodes in the multicast tree is as the value of fitness function;

Chromosome C _jThe multicast tree of the correspondence of trying to achieve based on shortest path first is designated as T (j), and all maximums to the path delay of destination node are designated as among the multicast tree T (j):

$\mathrm{MD}\left(C_j\right)=\underset{r_i\∈R}{\mathrm{Max}}\left\{\mathrm{Delay}\left(P_{T\left(j\right)}(S,r_i)\right)\right\}$ j∈{1，2，...N}；

Chromosome C _jFitness function value computing formula be:

F(C _j)＝1/MD(C _j)；

4) select operation:

According to the chromosomal adaptive value that fitness function in the step 3) calculates, adopt strategy according to qualifications, the highest individuality of ideal adaptation degree value is directly remained in the progeny population, according to each chromosomal ideal adaptation degree value, be calculated as follows out the relative adaptation value again: $p\left(C_j\right)=\frac{F\left(C_j\right)}{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}F\left(C_i\right)};$

Wherein, p (C _j) be this chromosomal selection probability, F (C _j) the ideal adaptation degree value of expression chromosome j, N is population size, the population size number is chromosome number;

5) interlace operation:

I) select two chromosome C at random _iAnd C _j

Ii) judge two chromosome C _iAnd C _jWhether different, if, then change step I ii, otherwise, step I changeed;

Iii) select an intermediate node v at random _aAs the crosspoint, v _aBe one arrive all destination nodes in the set of multicast destination node must through intermediate node v _a

Iv) multicast destination node counter k=0;

V) with chromosome C _iThe middle destination node r that arrives _kSubpath With chromosome C _jThe middle destination node r that arrives _kSubpath

Intersect;

Vi) destination node counter k adds 1, and intersection arrives the subpath of next destination node;

Vii) judge whether to have operated the path that arrives all destination nodes, if, finish, otherwise, step v changeed;

6) mutation operation:

I) selective staining body C at random _jCarry out mutation operation;

Ii) selective staining body C at random _jIn i bar limit carry out mutation operation;

Iii) i bar limit is produced a new channel NewC at random, channel NewC must be two node v that are connected with i bar limit _aAnd v _bThe channel that all has;

Iv) judge NewC whether with c _i(j) do not wait, if, finish, otherwise, step I i changeed;

7) produce population of future generation;

8) the genetic algebra counter adds 1, carries out follow-on genetic manipulation;

9) whether judge genetic algebra greater than GEN, if, change step 10, otherwise, step 3 changeed;

10) output optimal case:

To have the chromosome of maximum adaptation degree value as optimal case, in this chromosome, the maximum in the path delay from source node to all destination nodes is littler than the maximum in the path delay from source node to all destination nodes other chromosomes.

In the step 4, for p (C _j) be selected greater than the chromosome of given threshold xi=0.2 and enter the next generation, then be eliminated less than the chromosome of ξ, carry out interlace operation in the step 5, crossover probability p _c=0.6; In the step 6, the variation Probability p _m=0.05.

Beneficial effect:

The invention solves the multicast path of single source node in the cognitive radio Mesh network by reaching the spectrum allocation may problem, and can realize that source node arrives the maximum minimum in the path delay of all multicast destination nodes, and the average retardation to all destination nodes is also minimum, effectively raises network performance and the availability of frequency spectrum.

The present invention is described in further detail below in conjunction with accompanying drawing.

Description of drawings

Fig. 1 is the schematic diagram of cognitive radio Mesh network model of the present invention;

Fig. 2 is the schematic diagram of the potential interference relationships of the present invention;

Fig. 3 is that multicast path of the present invention is by the flow chart that reaches frequency spectrum distributing method;

Fig. 4 is the flow chart of initialization population of the present invention;

Fig. 5 the present invention is based on the chromosome C that shortest path first is tried to achieve _iAnd C _jThe schematic diagram of corresponding multicast tree;

Fig. 6 is the flow chart of interlace operation of the present invention;

Fig. 7 is the flow chart of mutation operation of the present invention.

Wherein 1 is chromosome C _iBased on the corresponding multicast tree T (i) that shortest path first is tried to achieve, 2 is chromosome C _jThe multicast tree T (j) of the correspondence of trying to achieve based on shortest path first.

Embodiment

Below with reference to the drawings and specific embodiments the present invention is described in further details:

Embodiment 1:

In the present embodiment, the set K={1 of available channel, 2,3,4,5,6}.The cognitive radio Mesh network that static CR-Mesh router is formed is modeled as a undirected multigraph G=, and (D), wherein V represents the set of CR-Mesh router for V, E.E represents to link the set of the Radio Link of two CR-Mesh routers that can intercom mutually, the meaning that two CR-Mesh routers can intercom mutually is that two CR-Mesh routers must have an identical available channel at least, and satisfy the constraint of communication distance, when two CR-Mesh routers have a plurality of identical available channel, there are many limits between them, be multi wireless links, so G is the non-directed graph that has multiple limit.D represents the set of the transmission delay of Radio Link between two CR-Mesh router nodes, and the transmission delay of different channels is different, and the transmission delay of supposing channel i is d _iEach node v _i∈ V has the set of available channels K of a perception _iEach node v _iAll there is a communication distance d in ∈ V _TrWith an interference distance d _Ir3d is arranged generally speaking _Tr＞d _Ir＞d _Tr, d of the present invention _Ir=2*d _Trd _IjExpression node v _iWith node v _jBetween physical distance.All nodes adopt half-duplex mode work.

Ψ _{I, j}Node is represented v _iWith node v _jIdentical set of available channels.S represents the multicast source node.R={r ₁, r ₂... r _mExpression multicast destination node.G '=(V ', E ', D ') represent the simple and connected graph by the G derivation, wherein, V '=V,

Any two nodes only have a usable wireless link among the G '.T=(V _T, E _T) the expression multicast tree, P _T(S, r _i) among expression multicast tree (chromosome) T one from S to r _iThe path.d _LDelay among the expression multicast tree T on the Radio Link L, d _LValue be to be determined by the channel that Radio Link L distributes because different channels has different delays.Delay (P _T(S, r _i)) among the expression multicast tree T from source node S to destination node r _iPath delay, namely

As shown in Figure 1,10 PU that distributing, 15 CR-Mesh routers.6 available channels are arranged, i.e. K={1,2,3,4,5,6} in the system.The transmission delay of 6 available channels is as shown in the table:

Channel	1	2	3	4	5	6
							Postpone (ms)	10	15	19	5	8	3

The channel that main user PU1 takies is 3, the channel that main user PU2 takies is 4, the channel that main user PU3 takies is 6, the channel that main user PU4 takies is 6, and the channel that main user PU5 takies is 4, and the channel that main user PU6 takies is 5, the channel that main user PU7 takies is 5, the channel that main user PU8 takies is 1, and the channel that main user PU9 takies is 5, and the channel that main user PU10 takies is 3.CR-Mesh route (CR-MR) available channel situation is as shown in the table:

CR_MR#	K i	CR_MR#	K i
				CR_MR1	{1，2，5}	CR_MR9	{1，2，3，4}
CR_MR2	{1，2，4，5，6}	CR_MR10	{2，3，6}
				CR_MR3	{1，2，3，5}	CR_MR11	{1，2，3，4，6}
CR_MR4	{2，4，5}	CR_MR12	{1，2，4，6}
				CR_MR5	{1，2，3}	CR_MR13	{2，3，4，6}
CR_MR6	{2，3，5，6}	CR_MR14	{1，2，4，6}
				CR_MR7	{1，2，3}	CR_MR15	{1，2，4，5，6}
CR_MR8	{1，2，3，4}

The potential interference relationships figure of network topological diagram shown in Figure 1 as shown in Figure 2, the channel that this Radio Link of the value representation on the limit can be used, as being 2/5 on the limit between CR-MR1, the CR-MR4 node, available on expression channel 2 and channel 5 these limits.As can be seen from Figure 2, Radio Link between CR-MR1, the CR-MR4 node, and there is potential interference relationships in the Radio Link between CR-MR4, the CR-MR7 node at channel 2, if namely CR-MR1 uses channel 2 to the Radio Link between the CR-MR4, then CR-MR4 just can not use channel 2 to the Radio Link between the CR-MR7.If earlier to CR-MR1 to the wireless link distribution channel 2 between the CR-MR4, then do not disturbing CR-MR1 under the situation of Radio Link between the CR-MR4, CR-MR4 will not have the channel that can distribute to the Radio Link between the CR-MR7.

The spectrum allocation may problem has been proved to be NP completely, the associating multicast path comprises the spectrum allocation may problem by the problem that reaches spectrum allocation may, therefore multicast path by and the problem of spectrum allocation may also be NP completely, we adopt the genetic algorithm for solving multicast path by and the approximate solution of spectrum allocation may problem.Be illustrated in figure 3 as the multicast path of the present invention's proposition by the flow chart that reaches frequency spectrum distributing method, wherein GEN represents genetic algebra, is that 200. steps are as follows in the present embodiment:

S1-1 initialization genetic algebra counter g=0.

S1-2 initialization population, at first we set chromosomal coded representation mode, use length as n (n=|E ' |, the limit collection of the simple and connected graph G ' that E ' presentation graphs G derives, by n=23 as shown in Figure 2) 7 system string lists show a chromosome, representing a channel assignment scheme, we give, and n bar limit is numbered from 1 to n among the multicast tree T, and coded representation is X=x ₁x ₂X _n, x _i∈ 0}U K, i ∈ 1,2 ..., n} is in chromosome, if x _i=0, any channel is not distributed on the limit that is numbered i in the expression multicast tree, and so, at this moment, the delay on the Radio Link that this limit is represented will be infinity, if x _i=k, k ∈ K illustrates that then the channel of the limit distribution that is numbered i is k.

Determined that we generate initial population at random, as shown in Figure 4 after the chromosomal coded representation mode.We suppose that population number is N, N=80, and chromosome j is expressed as C in the population _j=c ₁(j) c ₂(j) ... c _n(j), wherein n is the limit number of simple and connected graph G ', c _i(j) ∈ 0}U K, i ∈ 1,2 ..., n}, j ∈ 1,2 ..., N}, N are the chromosome sum in the population.The initialization population comprises following steps:

S2-1 initialization chromosome counting device j=0.

S2-2 chromosome C _jBe initialized as sky.

S2-3 initialization chromosome C _jLimit counter i=0.

S2-4 initialization chromosome C _jC _i(j) channel on limit is supposed chromosome C _jTwo nodes connecting of i bar limit be v _aAnd v _b, c then _i(j) value of Fen Peiing necessarily belongs to Ψ _{A, b}Set namely must distribution node v _aAnd v _bAvailable channels all, c _i(j) be Ψ _{A, b}In any one value, if Ψ _{A, b}=Φ, then c _i(j)=0, i.e. hinged node v _aAnd v _bI bar limit do not distribute any channel.

S2-5 limit counter adds 1, and next jumps the channel that the limit is distributed the expression initialization.

S2-6 judges i 〉=n, and whether all limits of initialization if change S2-7; Otherwise, change S2-4.

S2-7 chromosome counting device adds 1, the next chromosome of expression initialization.

S2-8 judges j 〉=N, and whether all chromosome of initialization if finish; Otherwise, change S2-2.

S1-3 calculates ideal adaptation degree value based on shortest path first.

In genetic algorithm, the fitness function value is to weigh the standard of chromosome quality in the population, and it will directly reflect chromosomal performance, according to the size of fitness function value, determines some chromosome to be breeding or to wither away.

The present invention is based on shortest path first and find the solution source node S to the multicast tree of destination node set R, the delay of Radio Link is designated as the weights on limit.Ask source node S to gather the shortest path of all destination nodes among the R to destination node.The peaked inverse in source node S path delay of all destination nodes in the multicast tree is as the value of fitness function.If chromosome C _jThe multicast tree of the correspondence of trying to achieve based on shortest path first is designated as T (j), and all maximums to the path delay of destination node are designated as among the multicast tree T (j): $\mathrm{MD}\left(C_j\right)=\underset{r_i\∈R}{\mathrm{Max}}\left\{\mathrm{Delay}\left(P_{T\left(j\right)}(s,r_i)\right)\right\}$ j∈{1，2，...N}

Chromosome C _jFitness function value computing formula be:

F(C _j)＝1/MD(C _j)

Chromosomal fitness value is more big in N the chromosome, and we think that its performance is more good, and fitness value is more little, and performance is more poor.The maximum of the shortest path of all destination nodes among the set R will determine chromosomal performance from the source node S to the destination node.

Figure 5 shows that chromosome C _iAnd C _jTwo multicast tree T (i) and the T (j) of the correspondence of trying to achieve based on shortest path first.Suppose that CR-MR1 is source node S, CR-MR13, CR-MR14, CR-MR15 are multicast destination node set R.Shown in multicast tree T (i), be 5 (8) on the limit between CR-MR1 and the CR-MR5, the meaning is that the Radio Link between them uses channel 5, and transmission delay is 8.Source node CR-MR1 to the shortest path length of CR-MR13, CR-MR14, CR-MR15 in T (i) is, 45=8+15+19+3,48=8+15+10+15,41=8+15+10+5+3, in T (j) be, 63=10+19+15+19,47=10+19+15+3,49=10+15+19+5) as shown in the table:

Shortest path length	T(i)	T(j)
			Delay(P T(S，r CR-MR13))	45	63
Delay(P T(S，r CR-MR14))	48	47
			Delay(P T(S，r CR-MR15))	41	49

As seen from the above table, F (C _i)=1/48, F (C _j)=1/63, chromosome C _iHave bigger fitness value, show chromosome C _iPerformance more good, compare chromosome C _jHaving higher probability is selected and enters the next generation.

S1--4 selects operation.Select the main strategy that adopts the fitness value ratio of operation among the present invention.At first the chromosomal adaptive value that calculates according to fitness function adopts strategy according to qualifications then, and the individuality that adaptive value is the highest directly remains in the progeny population.According to each chromosomal adaptive value, be calculated as follows out its relative adaptation value at last:

$p\left(C_j\right)=\frac{F\left(C_j\right)}{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}F\left(C_i\right)}$

P (C _j) as this chromosomal selection probability, F (C _j) adaptive value of expression chromosome j, N is population size.

The S1-5 interlace operation.The present invention carries out interlace operation and at first selects two chromosome C at random from the population set _iAnd C _j, select at random all destination nodes in the set of some arrival multicast destination nodes must through intermediate node v _a, with node v _aFor carrying out interlace operation in the crosspoint.At chromosome C _iIn, arrive destination node r _kThe path

Can regard as

With

Equally, at chromosome C _jIn, arrive destination node r _kThe path

Can regard as

With The interlace operation flow process may further comprise the steps as shown in Figure 6:

S3-1 selects two chromosome C at random _iAnd C _j

S3-2 judges two chromosome C _iAnd C _jWhether different, if, then change S3-3, otherwise, S3-1 changeed.

S3-3 selects an intermediate node v at random _aAs the crosspoint.

S3-4 multicast destination node counter k=0.

S3-5 is with chromosome C _iThe middle destination node r that arrives _kSubpath

With chromosome C _jThe middle destination node r that arrives _kSubpath

Intersect.

S3-6 destination node counter k adds 1, and intersection arrives the subpath of next destination node.

S3-7 judges whether to have operated the path that arrives all destination nodes, if, finish, otherwise, S3-5 changeed.

The S1-6 mutation operation.The present invention carries out mutation operation selective staining body C at random _jCarrying out mutation operation, mainly is to chromosome C _jIn the channel value that distributes of certain bar limit make amendment.Chromosome C _jBe expressed as C _j=c ₁(j) c ₂(j) ... c _n(j), two nodes of i bar limit connection are v _aAnd v _b, Ψ _{A, b}Be node v _aWith node v _bIdentical set of available channels.The mutation operation flow process comprises following steps as shown in Figure 7:

S4-1 is selective staining body C at random _jCarry out mutation operation.

S4-2 is selective staining body C at random _jIn i bar limit carry out mutation operation.

S4-3 produces a new channel NewC at random to i bar limit, and NewC is two node v that i bar limit connects _aAnd v _bThe channel that all has.

S4-4 judge NewC whether with c _i(j) do not wait, if, finish, otherwise, S4-2 changeed.

S1-7 produces population of future generation.Select operation among the S1-4, for p (C _j) be selected greater than the chromosome of given threshold xi=0.2 and enter the next generation, then be eliminated less than the chromosome of ξ.Carry out interlace operation among the S1-5, crossover probability p _c=0.6.Carry out mutation operation among the S1-6, the variation Probability p _m=0.05.

S1-8 genetic algebra counter adds 1, carries out follow-on genetic manipulation.

Whether S1-9 judges genetic algebra greater than GEN=200, if change S1-10; Otherwise, change S1-3.

S1-10 exports optimal case.The chromosome that namely has maximum adaptation degree value will be as the anti-case of optimum.In this chromosome, the maximum in the path delay from source node to all destination nodes is littler than the maximum in the path delay from source node to all destination nodes other chromosomes.

For this example, optimal case is as follows: the routed path from source node CR-MR1 to CR-MR13 and channel allocation situation are: CR-MR1 is 5 to the channel that CR-MR4 distributes, CR-MR4 is 4 to the channel that CR-MR2 distributes, the channel that CR-MR2 distributes to CR-MR6 is that 5, CR-MR6 is 6 to the channel that CR-MR13 distributes; CR-MR1 to routed path and the channel allocation situation of CR-MR14 is: the channel that CR-MR1 distributes to CR-MR5 is that the channel that 1, CR-MR5 distributes to CR-MR8 is that the channel that 2, CR-MR8 distributes to CR-MR11 is that 4, CR-MR11 is 6 to the channel that CR-MR14 distributes; CR-MR1 to routed path and the channel allocation situation of CR-MR15 is: CR-MR1 is 1 to the channel that CR-MR5 distributes, CR-MR5 is 2 to the channel that CR-MR8 distributes, CR-MR8 is 4 to the channel that CR-MR11 distributes, the channel that CR-MR11 distributes to CR-MR12 is that 6, CR-MR12 is 4 to the channel that CR-MR15 distributes.

Claims (2)

1. Multicast path in the cognitive radio Mesh network by and frequency spectrum distributing method, it is characterized in that, at first the cognitive radio Mesh network that static CR-Mesh router is formed is modeled as a undirected multigraph G=(V, E, D), wherein V represents the set of CR-Mesh router, and E represents to link the set of the Radio Link of two CR-Mesh routers that can intercom mutually, D represents the set of the transmission delay of Radio Link between two CR-Mesh router nodes, d _iThe transmission delay of representing a certain channel i, each node v _i∈ V has the set of available channels K of a perception _i, all nodes adopt half-duplex mode work;

   K={1,2 ... k}={1,2,3,4,5,6} represent the set of total available channel, Ψ _{I, j}Expression node v _iWith node v _jIdentical set of available channels, S represents multicast source node, R={r ₁, r ₂... r _mThe set of expression multicast destination node; G'=(V', E', the D') simple and connected graph that derived by G of expression, wherein, V'=V,

   

   Any two nodes only have a usable wireless link among the G';

   T=(V _T, E _T) the expression multicast tree, P _T(S, r _i) among the expression multicast tree T one from S to a certain destination node r _iThe path, d _LDelay among the expression multicast tree T on the Radio Link L, Delay (P _T(S, r _i)) among the expression multicast tree T from source node S to destination node r _iPath delay, namely

   

   Take genetic algorithm to ask for the channel assignment scheme of the maximum minimum in path delay then, GEN represents genetic algebra, and step is as follows:

   1) initialization genetic algebra counter g=0;

   2) initialization population, at first set chromosomal coded representation mode, use length to show a chromosome as the 7 system string lists of n, representing a channel assignment scheme, n=|E'|=23 wherein, the limit number of the simple and connected graph G' that E' presentation graphs G derives is numbered from 1 to n to n bar limit among the multicast tree T, and coded representation is X=x ₁x ₂X _n, x _i∈ 0} ∪ K, i ∈ 1,2 ..., n} is in chromosome, if x _i=0, any channel is not distributed on the limit that is numbered i in the expression multicast tree, and the delay on the Radio Link that this limit is represented will be infinity, if x _i=k, k ∈ K illustrates that then the channel of the limit distribution that is numbered i is k;

   Chromosome in the population adds up to N, and chromosome j is expressed as C in the population _j=c ₁(j) c ₂(j) ... c _n(j), c _i(j) ∈ 0} ∪ K, i ∈ 1,2 ..., n}, j ∈ 1,2 ..., N}, the initialization population comprises following steps:

   ⅰ) initialization chromosome counting device j=0;

   ⅱ) chromosome C _jBe initialized as sky, j ∈ 1,2 ..., N};

   ⅲ) initialization chromosome C _jLimit counter i=0;

   ⅳ) initialization chromosome C _jC _i(j) channel on limit is set chromosome C _jTwo nodes connecting of i bar limit be v _aAnd v _b, initialization c then _i(j) for belonging to Ψ _{A, b}If any value in the set is Ψ _{A, b}=Φ, then c _i(j)=0, i.e. connected node v _aAnd v _bI bar limit do not distribute any channel;

   ⅴ) the limit counter adds 1, the channel that next bar limit of expression initialization is distributed;

   ⅵ) size of judgement i and n determines whether the limit that initialization is all, if change step ⅶ; Otherwise, change step ⅳ;

   ⅶ) the chromosome counting device adds 1, the next chromosome of expression initialization;

   ⅷ) size of judgement j and N determines whether the chromosome that initialization is all, if finish; Otherwise, change step ⅱ;

   3) calculating ideal adaptation degree value based on shortest path first is the fitness function value:

   Find the solution source node S to the multicast tree of destination node set R based on shortest path first, the delay of Radio Link is designated as the weights on limit, ask the shortest path of source node S all destination nodes in the destination node set R, the peaked inverse in source node S path delay of all destination nodes in the multicast tree is as the value of fitness function;

   Chromosome C _jThe multicast tree of the correspondence of trying to achieve based on shortest path first is designated as T (j), and all maximums to the path delay of destination node are designated as among the multicast tree T (j):

   $\mathrm{MD}\left(C_j\right)=\underset{r_i\∈R}{\mathrm{Max}}\left\{\mathrm{Delay}\left(P_{T\left(j\right)}(S,r_i)\right)\right\}j\∈\{1,2,...N\};$

   Chromosome C _jFitness function value computing formula be:

   F(C _j)=1/MD(C _j)；

   4) select operation:

   According to the chromosomal adaptive value that fitness function in the step 3) calculates, adopt strategy according to qualifications, the highest individuality of ideal adaptation degree value is directly remained in the progeny population, according to each chromosomal ideal adaptation degree value, be calculated as follows out the relative adaptation value again:

   

   Wherein, p (C _j) be this chromosomal selection probability, F (C _j) the ideal adaptation degree value of expression chromosome j, N is population size, the population size number is chromosome number;

   5) interlace operation:

   ⅰ) select two chromosome C at random _iAnd C _j

   ⅱ) judge two chromosome C _iAnd C _jWhether different, if, then change step ⅲ, otherwise, step ⅰ changeed;

   ⅲ) select an intermediate node v at random _aAs the crosspoint, v _aBe one arrive all destination nodes in the set of multicast destination node must through intermediate node v _a

   ⅳ) multicast destination node counter k=0;

   ⅴ) with chromosome C _iThe middle destination node r that arrives _kSubpath

   

   With chromosome C _jThe middle destination node r that arrives _kSubpath

   

   Intersect;

   ⅵ) destination node counter k adds 1, and intersection arrives the subpath of next destination node;

   ⅶ) judge whether to have operated the path that arrives all destination nodes, if, finish, otherwise, step ⅴ changeed;

   6) mutation operation:

   ⅰ) selective staining body C at random _jCarry out mutation operation;

   ⅱ) selective staining body C at random _jIn i bar limit carry out mutation operation;

   ⅲ) i bar limit is produced a new channel NewC at random, channel NewC must be two node v that are connected with i bar limit _aAnd v _bThe channel that all has;

   ⅳ) judge NewC whether with c _i(j) do not wait, if, finish, otherwise, step ⅱ changeed;

   7) produce population of future generation;

   8) the genetic algebra counter adds 1, carries out follow-on genetic manipulation;

   9) whether judge genetic algebra greater than GEN, if, change step 10, otherwise, step 3 changeed;

   10) output optimal case:

   To have the chromosome of maximum adaptation degree value as optimal case, in this chromosome, the maximum in the path delay from source node to all destination nodes is littler than the maximum in the path delay from source node to all destination nodes other chromosomes.
2. 2. the multicast path in a kind of cognitive radio Mesh network according to claim 1 is characterized in that, in the step 4, for p (C by reaching frequency spectrum distributing method _j) be selected greater than the chromosome of given threshold xi=0.2 and enter the next generation, then be eliminated less than the chromosome of ξ, carry out interlace operation in the step 5, crossover probability p _c=0.6; In the step 6, the variation Probability p _m=0.05.

Images