CN103685020A - Genetic algorithm based minimum-cost optical multicast tree routing method - Google Patents
Genetic algorithm based minimum-cost optical multicast tree routing method Download PDFInfo
- Publication number
- CN103685020A CN103685020A CN201310606366.9A CN201310606366A CN103685020A CN 103685020 A CN103685020 A CN 103685020A CN 201310606366 A CN201310606366 A CN 201310606366A CN 103685020 A CN103685020 A CN 103685020A
- Authority
- CN
- China
- Prior art keywords
- chromosome
- gene
- fitness function
- destination node
- multicast tree
- 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.)
- Granted
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a genetic algorithm based minimum-cost optical multicast tree routing method. The method includes two parts of edge initialization of a network and minimum-cost multicast tree iteration. The edge initialization of the network mainly includes completing initialization of edges in the network, representing the edges with integer multiples of unit capacity with multiple unit capacity edges, and conveniently applying a genetic algorithm to optimize an information transmission path and an encoding method. The minimum-cost optical multicast tree iteration mainly includes the steps of selection, intersection, variation, bad gene removal and the like. Some bad genes are removed from a gene pool according to a designed fitness function value in the process of iteration every time, so that size of an algorithm searching space can be greatly decreased, convergence rate of the algorithm is increased beneficially, and an optical multicast tree with less cost is found out. The information transmission routing method is smallest in total number of found and needed information transmission links meeting the requirement on multicast request rate, and minimum in encoding operation times.
Description
Technical field
The present invention relates to communication technical field, be specifically related to a kind of light multicast tree minimum cost method for routing based on genetic algorithm.
Background technology
Fast development along with multicast services such as video conference, Web TV, multimedia remote educations, the quick increase of the consumption of the network bandwidth and congested generation, traditional communication net is being faced with the serious problems such as bandwidth resources are not enough, network throughput is low, service blocking rate rising, and the resource day of network is becoming tight.Because optical-fiber network has the message transmission capability of high bandwidth and two-forty, it has temporarily solved the problem such as bandwidth resources deficiency in the network of traditional electrical territory.But, in recent years along with the maintaining sustained and rapid growth of broadband services and multicast application, optical-fiber network faces the problems such as bandwidth resources deficiency, network blocking probability continue to increase again.Network code is improved network throughput, balance network load, increase network bandwidth utilization factor, reduces network resource loss, improves internet security, reduces the advantages such as energy consumption, thereby the routing mode of coding Network Based is more and more subject to researcher's attention.Network code is incorporated in optical-fiber network, utilizes the advantage of network code to solve the problems such as the optical network band width inadequate resource effective method of can yet be regarded as.Therefore, we propose the building method solution minimum cost light multicast tree routing problem of many places, a kind of single source minimum cost multicast tree.
Be different from traditional routing mode, the routing mode of coding Network Based not only allows the intermediate node of network that the information of receiving is stored and forwarded, and can also carry out information coding squeeze operation.Code construction light multicast tree route Network Based generally comprises 2 steps: (1) determines the information receiving velocity of network destination node; (2) according to the information receiving velocity of destination node, determine limit disjoint paths number for each destination node.For make network destination node correct decode raw information, all destination node receiving velocities are identical, this speed is generally less than the minimum value in the max-flow that equals all destination nodes.But, at the intermediate node of communication network, carry out network code operation, will increase the corresponding expense of communication network and cost, as: the complexity that increases intermediate node computes; Increase the demand of buffer, for storing the information on decoding input limit; Bring network delay increase etc.Thereby the application's minimum cost multicast tree is to find to meet that the required link number of light multicast route summation is minimum, a kind of information transferring method of encoding operation least number of times.
Studies have shown that now: minimum cost light multicast tree is a NP-complete problem, current solution major part is to adopt heuritic approach to solve this problem.But, existing heuritic approach generally in particular network effect better, and effect is undesirable in other network, and intelligent optimization algorithm is when solving NP-complete problem, effect will be better than heuritic approach, and genetic algorithm is a kind of algorithm comparatively ripe in intelligent optimization algorithm.Therefore, we propose a kind of light multicast tree minimum cost method for routing based on genetic algorithm.
Summary of the invention
For above deficiency of the prior art, the object of the present invention is to provide a kind of bandwidth resources utilance that can improve significantly optical-fiber network, reduce the light multicast tree minimum cost method for routing based on genetic algorithm of network code number of operations.Technical scheme of the present invention is as follows: a kind of light multicast tree minimum cost method for routing based on genetic algorithm, and it comprises the following steps:
101, obtain network topology G(V, E), wherein V represents the set of node of network topology G, E represents the fillet between nodes, when capacity n >=2 of fillet, this fillet is converted into n bar side by side and limit that capacity is 1, complete initialization, jump to step 102;
102, in obtaining step 101 after initialization network topology G(V, E) source node S and destination node collection t, structure source node S is to the multicast tree of destination node collection t, determine that source node S is to the maximum multicast speed T of destination node collection t, and the receiving velocity of setting destination node is k, wherein 1≤k≤T; Obtain source node S to destination node t
ithe N paths of all existence, wherein destination node t
ian element in destination node collection t, calculates this destination node t
ithe mode m of k bar limit combination of paths
iproduce gene pool, and adopt genetic algorithm structure source node S to the chromosome population of destination node collection t, each chromosome represents a kind of routing mode of network, wherein each chromosome is by U the genomic constitution equating with the number of destination node, and each gene representation source node S is to corresponding destination node t
ia kind of path;
103, chromosomal fitness function in constitution step 102
F=a
1* NC (R)+a
2* N
cL, and a
1>a
2
In formula, f is fitness function value; NC (R) is for meeting the link cost of the multicast tree of multicast request rate, N
cLfor coding strand way order; a
1, a
2for weight coefficient;
When fitness function f is greater than or equal to set point number N1 according to the number of times of genetic algorithm iteration renewal, export the chromosomal path of this optimum and fitness function value f, jump to step 106; Or the number of times upgrading according to genetic algorithm iteration as fitness function f is greater than N2 and fitness function value f when constant, export the chromosomal path of this optimum and fitness function value f, jump to step 106, finish; Otherwise, jump to step 104;
104, adoption rate back-and-forth method joins in initial chromosome population the chromosome in step 103, and through intersection step, variation step, tries to achieve optimum chromosome and fitness function value successively;
105, in the chromosome population of the optimum chromosome of trying to achieve in step 104 and the foundation of fitness function value substitution step 102, delete the gene that fitness function value is greater than this optimum chromosome fitness function value;
106, export final optimum chromosome and fitness value thereof, and to the path of destination node, carry out route according to the source node of this optimum chromosome representative.
Further, the ratio back-and-forth method in step 104 is that wheel disc is selected or Monte Carlo back-and-forth method.
Further, the intersection step in step 104 comprises:
A1, choose 2 chromosomes as parent chromosome at random;
A2, each gene in chromosome is produced to one 0 to 1 between random digit, for judging whether 2 parent chromosomes carry out interlace operation at random;
The numeral of A3, random generation in steps A 2 is less than p
ctime, 2 chromosomes carry out chiasma operation and produce 2 child chromosome, and whether the random numeral producing of determining step A2 is less than p
cif this random value is less than p
c, gene corresponding in 2 parent chromosomes is carried out to cross exchanged; Otherwise gene corresponding in 2 parent chromosomes remains unchanged; P wherein
cfor crossover probability, crossover probability p wherein
cspan is 0.6～0.98;
Whether the fitness function value of the child chromosome producing in A4, determining step A3 is less than the chromosomal fitness function value of parent, if so, jumps to steps A 5;
A5, by child chromosome, replace the larger parent chromosome of fitness function value in parent.
Further, the variation step in step 104 comprises:
B1, select a chromosome, each gene in chromosome is produced to one 0 to 1 between random digit;
In B2, determining step B1, whether the random digit of each gene is less than variation Probability p
m, if so, in gene pool corresponding to this destination node, integer of random selection replaces the numeral on original gene position, selects another kind of limit disjoint paths compound mode; Otherwise the data keeping in this gene position is constant.
Advantage of the present invention and beneficial effect are as follows:
The present invention is first in the initialized figure of network edge capacity, search source node is to all paths of each multicast destination node, and in these paths, find the compound mode of all limits disjoint paths meet multicast request rate, then with limit disjoint paths compound mode, construct chromosome, a kind of corresponding chromosomal gene of limit disjoint paths compound mode that meets multicast rate request, like this, a chromosome represents a light multicast tree, and genetic algorithm is selected, intersection, mutation operation be all based on chromosome; In order to prevent the evolution of genetic algorithm, has no object, thereby this patent is provided with according to minimum cost multicast tree optimization aim the evolution direction that corresponding fitness function is controlled light multicast tree, make that light multicast tree is minimum towards link number summation, the direction of encoding operation least number of times is evolved; In order to prevent that some bad genes are genetic in chromosome of future generation, this patent is by removing these genes inferior to reach the object of the cost that reduces required smooth multicast tree.
Accompanying drawing explanation
A kind of light multicast tree minimum cost method for routing flow chart based on genetic algorithm of Fig. 1;
The initialization of Fig. 2 network edge;
Fig. 3 gene pool make;
The operation of Fig. 4 chiasma;
The operation of Fig. 5 chromosomal variation.
Embodiment
The invention will be further elaborated below in conjunction with accompanying drawing, to provide the embodiment of an indefiniteness.
The present invention proposes a kind of light multicast tree minimum cost method for routing based on genetic algorithm.To arriving the light multicast request of optical-fiber network input node, first according to the limit of light multicast request rate initialization network, in network after the initialization of limit, find all compound modes that meet the limit disjoint paths of multicast rate request, structural gene storehouse, and construct the multicast tree chromosome that meets multicast request, chromosome is selected, intersects, makes a variation, removed the iterative process of gene inferior and find minimum cost light multicast tree.Wherein, removing gene inferior is whether the fitness function value of answering by calculating gene pairs determines higher than optimum chromosomal fitness function value in epicycle iteration whether this gene is gene inferior.
Below in conjunction with accompanying drawing, technical scheme of the present invention is described further.
As shown in Figure 1, its concrete implementation step is as follows for a kind of flow chart of the light multicast tree minimum cost method for routing based on genetic algorithm:
The 1st step: read network topology G(V, E), in network, edge capacity is all the integral multiple of unit discharge, and the limit that is n (n >=2) by edge capacity in network is converted into the limit that n bar capacity is 1, i.e. unit discharge limit completes the initialization procedure of network edge.
The 2nd step: determine the information receiving velocity of multicast destination node, structure chromosome population.First according to network topology and multicast request, determine the receiving velocity of destination node, be assumed to be k (1≤k≤T, T is maximum multicast speed); Then, find source node to all paths of each destination node, for each destination node, determine the mode that k bar limit disjoint paths combines, hypothesis goal node i has m
iplant compound mode, this destination node has m
iplant gene.In the present invention, each chromosome is by t genomic constitution, and the number that wherein t is destination node, at random from [1, m
i] in choose an integer as the value of gene i.
The 3rd step: whether evaluation algorithm meets end condition.Here end condition is 2, satisfied wherein any one just exports optimum chromosome and fitness function value: the number of times of (1) iteration reaches and presets times N 1, presetting number of times is to determine according to network size size, as 14 meshed networks can arrange iteration 200 times; (2) in 2 generations of optimum chromosomal fitness value N continuous, all do not change, here the preferred value 20 of N2.Meet above-mentioned one of them condition, jump to the 9th step.
The 4th step: chromosome is selected.The effect that chromosome is selected is that the chromosome of choosing a part of fitness function value≤N3 from the 2nd step in the chromosome population producing joins in initial chromosome population, here our adoption rate system of selection selective staining body of the preferred value of N3, ratio back-and-forth method is also known as wheel disc and selects or Monte Carlo back-and-forth method.The basic thought of ratio back-and-forth method is to allow fitness function value (here for f
j) the selected probability of less individuality is larger, the selected probability of individuality that fitness function value is larger is less.If the selecteed probability of chromosome j is p
j, its expression formula is:
In above formula, N is chromosomal number, f
ifor the fitness function value of chromosome i, all chromosome fitness function values
a determined value, so, molecule f
jless p
jjust larger, the selected probability of chromosome j is just larger.
The 5th step: chiasma.Chiasma operates in the effect of playing a genetic recombination in genetic algorithm, and biology is constantly evolved and played vital effect, its essence is the gene of parent is exchanged with certain rule, and produce new individuality, specific operation process as shown in Figure 4.
The 6th step: chromosomal variation.Chromosomal variation is to prevent from being absorbed in locally optimal solution in genetic algorithm iterative process, causes finding the multicast tree of minimum cost, and chromosomal variation process as shown in Figure 5.
The 7th step: store optimum chromosome and fitness function value, preserve the minimum cost light multicast tree finding in epicycle iteration.In order effectively to search out, to meet the required link number of light multicast request rate summation minimum, a kind of information transmission mode of encoding operation least number of times, the design of fitness function is just particularly crucial, and according to the application's optimization aim, we design the cost that a kind of fitness function is evaluated multicast tree:
F=a
1* NC (R)+a
2* N
cL, and a
1>a
2
In formula, f is fitness function value; NC (R) is for meeting the link cost of a multicast tree of multicast request rate R; N
cLfor coding strand way order; a
1, a
2for the weight coefficient arranging, and a
1be greater than a
2thereby, make a
1* NC (R) plays dominance effect to fitness function f.Like this, first algorithm will remove to find the multicast tree of total Least-cost, then, passes through on this basis a
2* N
cLcarry out the number of coding nodes in regulating networks, finally, search out the scheme of coding strand road minimum number.Such as, a
1value 10, a
2 value 1.
The 8th step: remove the gene inferior in gene pool.According to optimum chromosomal fitness function value in the 7th step, Gene sufficiency functional value in gene pool can be greater than to the gene elmination of optimum chromosome fitness function value, like this, can reduce the size in Genetic algorithm searching space, the convergence rate of accelerating algorithm, fast searching is to more excellent solution.After gene elmination inferior, algorithm jumps to the 3rd step;
The 9th step: export final optimum chromosome and fitness value thereof.
In a kind of light multicast tree minimum cost method for routing based on genetic algorithm, first need the limit initialization of the network G of integer edge capacity, network edge initialization procedure is as shown in Figure 2.
In annex map 2, in network diagram 2 (a), the capacity on the digitized representation figure limit on limit, the capacity in network diagram G of first finding out is more than or equal to 2 limit, and limit (3,5) are that capacity is 2 limit, and other is all that capacity is 1 limit; Then by being that 1 equality limit substitutes with 2 capacity between the father node 3 of limit (3,5) and child node 5, as shown in network diagram 2 (b).
In order to find minimum cost light multicast tree, need to build the chromosomal gene pool of formation, find the compound mode of all limits disjoint paths.Fig. 3 is the exemplary plot of gene pool structure, and Fig. 3 (a) is a network topological diagram, and in network, all limits capacity after initialization is all 1, and source node is s, destination node t
1and t
2, the gene pool construction process of network diagram 3 (a) is as follows:
The 1st step: in calculating chart 3 (a), the maximum multicast speed of destination node is that 3(source node s is to destination node t
1max-flow be 3, source node s is to destination node t
2max-flow be 3, so maximum multicast speed is min (3,3)), in order to be correctly decoded out reception information, destination node receiving velocity need to be less than or equal to 3, the receiving velocity that we establish destination node is here 2;
The 2nd step: determine that source node s is to destination node t
1the path likely existing, as shown in Fig. 3 (b), comprises 7 paths, and to all path digital numberings; Same method, finds source node s to destination node t
2the path likely existing, as shown in Fig. 3 (c), comprises 7 paths, and to this 7 paths numeral number;
The 3rd step: be destination node t
1determine the mode of 2 limit disjoint paths combinations, structure destination node t
1gene pool, as shown in Fig. 3 (d), comprise 9 kinds of compound modes (representing 9 kinds of values of this destination node gene), digitized representation in combination numbering path; Same method is destination node t
2the mode of finding 2 limit disjoint paths combinations, as shown in Fig. 3 (e), comprises 9 kinds of compound modes (representing 9 kinds of values of this destination node gene); The 4th step: due to the number of the corresponding destination node of chromosomal gene number, so chromosomal gene number is 2 in Fig. 3 (a).The 1st gene in chromosome (being 1 to 9 here) from the numeral shown in Fig. 3 (d) selected one at random; Chromosomal the 2nd gene (being 1 to 9 here) from the numeral shown in Fig. 3 (e) is selected one at random.For example, in chromosome (1,8), gene 1 represents destination node t
1the routing mode of choosing is the combination of (Isosorbide-5-Nitrae), that is: S → A → t
1and S → B → D → t
1; In like manner, gene 8 represents destination node t
2the routing mode of choosing is the combination of (4,5), that is: S → B → E → F → G → t
2and S → C → t
2, chromosome (1,8) is with regard to the routing mode shown in corresponding Fig. 3 (f).
By above-mentioned 4 steps, we can be for meeting a chromosome of business structure of multicast rate request in network.The chromosome having chosen is carried out to interlace operation and realize genetic recombination, produce new child chromosome, play and optimize chromosomal effect, promote multicast tree cost to reduce.Chiasma operating process as shown in Figure 4.The flow process of chiasma concrete operations is as shown in Fig. 4 (a), and process is:
The 1st step: choose at random 2 chromosomes as parent chromosome;
The 2nd step: random digit between each gene in chromosome is produced to 0 to 1, for judging whether 2 parent chromosomes carry out interlace operation at random;
The 3rd step: the random numeral producing is less than p in step (2)
ctime, 2 chromosomes carry out chiasma operation and produce 2 child chromosome, and child chromosome generation rule is: whether the random numeral producing of determining step (2) is less than p
cif this random value is less than p
c, gene corresponding in 2 parent chromosomes is carried out to cross exchanged, as the numeral that in Fig. 4 (b), the 3rd gene of number produces is from top to bottom less than p
c, thereby by the 3rd numeral that gene pairs is answered exchange of number from top to bottom in 2 parent chromosomes, in like manner the 7th gene of number also carried out digital exchange from top to bottom; If this random value is not less than p
c, gene corresponding in 2 parent chromosomes remains unchanged.P
cfor crossover probability.Crossover probability p
cvalue is wanted suitably, if crossover probability p
cbe worth too greatly, the excellent individual in chromosome population will rapidly disappear; If crossover probability p
cbe worth too littlely, can cause again search speed too slow.Therefore, general crossover probability p
cspan is 0.6～0.98, and we get p in this process
c=0.8.
The 4th step: whether the fitness function value of the child chromosome producing in determining step (3) is less than the chromosomal fitness function value of parent, if so, jumps to the 5th step; Otherwise algorithm finishes.
The 5th step: replace the larger parent chromosome of fitness function value in parent by child chromosome.
In order to prevent that chromosome is absorbed in local optimum in iteration, thereby cannot find minimum cost light multicast tree, need to carry out mutation operation to chromosome, increase chromosomal diversity, alternative light multicast tree is increased.As shown in Figure 5, concrete steps are chromosomal variation operating process:
The 1st step: select a chromosome, each gene in chromosome is produced to one 0 to 1 between random digit, for judging whether this gene makes a variation.
The 2nd step: in determining step (1), whether the random digit of each gene is less than p
m(p
mfor variation probability), if so, in gene pool corresponding to this destination node, integer of random selection replaces the numeral on original gene position, selects another kind of limit disjoint paths compound mode; Otherwise the data keeping in this gene position is constant.。Research shows: work as p
mwithin=0.1 o'clock, effect is relatively good, p
mfor variation probability; Thereby also get p here
m=0.1.
These embodiment are interpreted as only for the present invention is described, is not used in and limits the scope of the invention above.After having read the content of record of the present invention, technical staff can make various changes or modifications the present invention, and these equivalences change and modification falls into the light multicast tree minimum cost method for routing claim limited range that the present invention is based on genetic algorithm equally.
Claims (4)
1. the light multicast tree minimum cost method for routing based on genetic algorithm, is characterized in that, comprises the following steps:
101, obtain network topology G(V, E), wherein V represents the set of node of network topology G, E represents the fillet between nodes, when capacity n >=2 of fillet, this fillet is converted into n bar side by side and limit that capacity is 1, complete initialization, jump to step 102;
102, in obtaining step 101 after initialization network topology G(V, E) source node S and destination node collection t, structure source node S is to the multicast tree of destination node collection t, determine that source node S is to the maximum multicast speed T of destination node collection t, and the receiving velocity of setting destination node is k, wherein 1≤k≤T; Obtain source node S to destination node t
ithe N paths of all existence, wherein destination node t
ian element in destination node collection t, calculates this destination node t
ithe mode m of k bar limit combination of paths
iproduce gene pool, and adopt genetic algorithm structure source node S to the chromosome population of destination node collection t, each chromosome represents a kind of routing mode of network, wherein each chromosome is by U the genomic constitution equating with the number of destination node, and each gene representation source node S is to corresponding destination node t
ia kind of path;
103, chromosomal fitness function in constitution step 102
F=a
1* NC (R)+a
2* N
cL, and a
1>a
2
In formula, f is fitness function value; NC (R) is for meeting the link cost of the multicast tree of multicast request rate, N
cLfor coding strand way order; a
1, a
2for weight coefficient;
When fitness function f is greater than or equal to set point number N1 according to the number of times of genetic algorithm iteration renewal, export the chromosomal path of this optimum and fitness function value f, jump to step 106; Or the number of times upgrading according to genetic algorithm iteration as fitness function f is greater than N2 and fitness function value f when constant, export the chromosomal path of this optimum and fitness function value f, jump to step 106, finish; Otherwise, jump to step 104;
104, adoption rate back-and-forth method joins in initial chromosome population the chromosome in step 103, and through intersection step, variation step, tries to achieve optimum chromosome and fitness function value successively;
105, in the chromosome population of the optimum chromosome of trying to achieve in step 104 and the foundation of fitness function value substitution step 102, delete the gene that fitness function value is greater than this optimum chromosome fitness function value;
106, export final optimum chromosome and fitness value thereof, and to the path of destination node, carry out route according to the source node of this optimum chromosome representative.
2. a kind of light multicast tree minimum cost method for routing based on genetic algorithm according to claim 1, is characterized in that: the ratio back-and-forth method in step 104 is that wheel disc is selected or Monte Carlo back-and-forth method.
3. a kind of light multicast tree minimum cost method for routing based on genetic algorithm according to claim 1, is characterized in that, the intersection step in step 104 comprises:
A1, choose 2 chromosomes as parent chromosome at random;
A2, each gene in chromosome is produced to one 0 to 1 between random digit, for judging whether 2 parent chromosomes carry out interlace operation at random;
The numeral of A3, random generation in steps A 2 is less than p
ctime, 2 chromosomes carry out chiasma operation and produce 2 child chromosome, and whether the random numeral producing of determining step A2 is less than p
cif this random value is less than p
c, gene corresponding in 2 parent chromosomes is carried out to cross exchanged; Otherwise gene corresponding in 2 parent chromosomes remains unchanged; P wherein
cfor crossover probability, crossover probability p wherein
cspan is 0.6～0.98;
Whether the fitness function value of the child chromosome producing in A4, determining step A3 is less than the chromosomal fitness function value of parent, if so, jumps to steps A 5;
A5, by child chromosome, replace the larger parent chromosome of fitness function value in parent.
4. a kind of light multicast tree minimum cost method for routing based on genetic algorithm according to claim 1, is characterized in that, the variation step in step 104 comprises:
B1, select a chromosome, each gene in chromosome is produced to one 0 to 1 between random digit;
In B2, determining step B1, whether the random digit of each gene is less than variation Probability p
m, if so, in gene pool corresponding to this destination node, integer of random selection replaces the numeral on original gene position, selects another kind of limit disjoint paths compound mode; Otherwise the data keeping in this gene position is constant.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310606366.9A CN103685020B (en) | 2013-11-25 | 2013-11-25 | A kind of light multicast tree minimum cost method for routing based on genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310606366.9A CN103685020B (en) | 2013-11-25 | 2013-11-25 | A kind of light multicast tree minimum cost method for routing based on genetic algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103685020A true CN103685020A (en) | 2014-03-26 |
| CN103685020B CN103685020B (en) | 2017-07-28 |
Family
ID=50321380
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310606366.9A Active CN103685020B (en) | 2013-11-25 | 2013-11-25 | A kind of light multicast tree minimum cost method for routing based on genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103685020B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107294658A (en) * | 2017-07-03 | 2017-10-24 | 电子科技大学 | A kind of coding nodes choosing method in network control coding |
| CN107749819A (en) * | 2017-09-14 | 2018-03-02 | 北京东土科技股份有限公司 | Route selection method and device under the conditions of a kind of grid network |
| CN108400940A (en) * | 2018-02-27 | 2018-08-14 | 西南交通大学 | A kind of multicast virtual network function dispositions method based on Estimation of Distribution Algorithm |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102244840A (en) * | 2011-06-17 | 2011-11-16 | 中南大学 | Method for routing multicasts and allocating frequency spectrums in cognitive wireless Mesh network |
| CN102880186A (en) * | 2012-08-03 | 2013-01-16 | 北京理工大学 | Flight path planning method based on sparse A* algorithm and genetic algorithm |
-
2013
- 2013-11-25 CN CN201310606366.9A patent/CN103685020B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102244840A (en) * | 2011-06-17 | 2011-11-16 | 中南大学 | Method for routing multicasts and allocating frequency spectrums in cognitive wireless Mesh network |
| CN102880186A (en) * | 2012-08-03 | 2013-01-16 | 北京理工大学 | Flight path planning method based on sparse A* algorithm and genetic algorithm |
Non-Patent Citations (5)
| Title |
|---|
| REN-HUNG HWANG.ETC: "Multicast Routing Based on Genetic Algorithms", 《JOURNAL OF INFORMATION SCIENCE AND ENGINEERING》 * |
| 刘军等: "支持网络编码的认知无线自组网拓扑控制算法", 《通信学报》 * |
| 姜华: "遗传算法在QoS多播路由算法中的应用", 《中国优秀硕士论文电子期刊网》 * |
| 郑四海: "无线移动自组织网络QoS路由协议研究", 《中国博士学位论文电子期刊网》 * |
| 郝琨: "一种最小化编码节点的网络编码优化算法", 《电子与信息学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107294658A (en) * | 2017-07-03 | 2017-10-24 | 电子科技大学 | A kind of coding nodes choosing method in network control coding |
| CN107749819A (en) * | 2017-09-14 | 2018-03-02 | 北京东土科技股份有限公司 | Route selection method and device under the conditions of a kind of grid network |
| CN107749819B (en) * | 2017-09-14 | 2020-07-21 | 北京东土科技股份有限公司 | Routing method and device under grid network condition |
| CN108400940A (en) * | 2018-02-27 | 2018-08-14 | 西南交通大学 | A kind of multicast virtual network function dispositions method based on Estimation of Distribution Algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103685020B (en) | 2017-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cheng et al. | Joint QoS multicast routing and channel assignment in multiradio multichannel wireless mesh networks using intelligent computational methods | |
| CN101777990B (en) | Method for selecting multi-objective immune optimization multicast router path | |
| CN101677286B (en) | Optimization method of carrier network | |
| CN104853399B (en) | Cooperating relay system of selection based on improved Genetic Particle Swarm hybrid algorithm | |
| CN103701702A (en) | Dynamic routing algorithm in power line carrier communication | |
| CN105282815B (en) | A kind of more relay selection methods in DF collaborative networks | |
| CN103916220A (en) | Network coding cooperative communication method based on polarization codes | |
| CN101616074B (en) | Multicast routing optimization method based on quantum evolution | |
| CN102006221A (en) | Method for optimizing quality of service of stream media | |
| CN103685020A (en) | Genetic algorithm based minimum-cost optical multicast tree routing method | |
| CN105578555A (en) | Routing method of wireless ad hoc network, node apparatus and system | |
| CN102711125A (en) | Method for improving transmission capability of wireless mesh network | |
| CN103561457B (en) | A kind of multi-target networks power distribution method in heterogeneous wireless network collaboration communication | |
| CN103220086B (en) | High throughput network coding method in a kind of wireless network | |
| CN102148664B (en) | Inter-multicast network coding control method applied to multi-source multi-destination network | |
| CN104301259A (en) | Resource allocation method applicable to multi-hop wireless mesh network | |
| Huang et al. | Latency reduced method for modified successive cancellation decoding of polar codes | |
| CN108111991B (en) | D2D network building method based on scalable video streaming user experience quality | |
| CN101917334B (en) | Transmission network generation method by network coding of part of nodes | |
| CN115134928B (en) | Wireless Mesh network congestion control method with optimized frequency band route | |
| Chen et al. | A degree-delay-constrained genetic algorithm for multicast routing tree | |
| CN115208820A (en) | Automatic route planning method for RapidIO switching network | |
| CN105306374A (en) | QoS (Quality of Service) broadcast method for Overlay network based on genetic algorithm | |
| Huang et al. | A recursive partitioning algorithm for space information flow | |
| CN110932758B (en) | Low-voltage power line layered ant colony routing method based on non-overlapping clustering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |