Research On The Optimization Of Multicast Network Coding Based On Genetic Algorithm

With the vigorous development of the Internet,applications with high latency requirements,long maintenance time and huge amount of information are emerging.Multicast communication technology is the best communication method to support such applications.As a new data transmission technology,network coding can effectively improve the performance and security of multicast networks,and has attracted widespread attention.However,the operation of encoding and decoding in network coding results in additional consumption of network resources and aggravation of the complexity of network operations.This thesis focuses on the optimization of multicast network coding based on genetic algorithm,and the main work is as follows:This thesis analyzes the optimization problem of multicast network coding,and defines to minimize the coding node as the optimal objective and the mathematical model of optimization.To improve the defects of the existing SGA(Standard Genetic Algorithm)in solving network coding optimization problems,an optimization scheme for multicast network coding based on IA-GA(Improved Adaptive Genetic Algorithm)is proposed.On the one hand,this scheme preprocesses the initial network topology,filters out redundant links and nodes,and improves the efficiency of the algorithm.On the other hand,four improvements were made to SGA.The first is to optimize the initial population generation strategy so as to ensure the quality of the initial population;The second is to adopt adaptive crossover and mutation probability to reduce invalid genetic operations and speed up the convergence of the algorithm;The third is to use a mixed selection mechanism which combines roulette strategy and tournament strategy and takes into account the quality and diversity of genetic populations;The fourth is to design a new fitness function to accept inferior solutions with a certain probability to ensure the diversity of the population.In order to break through the limitations of genetic algorithm itself and further improve the optimization performance,this thesis introduces niche technology and tabu search algorithm on the basis of IA-GA,and proposes a method based on IAH-GA(Improved Adaptive Hybrid Genetic Algorithm)network coding optimization scheme.The introduction of niche technology can greatly enhance the diversity of genetic populations and avoid local problems;the introduction of tabu search algorithm can make up for the shortcomings of genetic algorithm’s low local search ability and accelerate the convergence speed.The schemes proposed in this thesis are compared on multiple sets of network topologies.The experimental results prove that the IA-GA scheme and the IAH-GA scheme are superior to the existing SGA algorithm scheme in terms of convergence speed,effectiveness and operating efficiency,and the IAH-GA scheme has more advantages in solving multicast network coding optimization than the IA-GA scheme.