Optimization Of Clonal Selection Algorithm And Research On Blind Equalization In MIMO Communication

Every aspect of biology has been a source of inspiration for developing new computational models and problem-solving methods.Traditional bionic algorithms Genetic Algorithm,Particle Swarm Algorithm,Ant Colony Algorithm)have the disadvantages of easy precocity and slow convergence.The immune system has become one of the most important areas because of its powerful information processing capabilities.The Immune Intelligence Algorithm(IIA)inspired by the immune system has the characteristics of fast convergence,less premature convergence,and excellent performance in solving optimization problems.The Clonal Selection Algorithm(CSA)is a new intelligent algorithm inspired by the biological immune system.It combines the prior knowledge of the problem with the two characteristics of the adaptive ability of the biological immune system,and thus has strong robustness in information processing,and can quickly converge to the global optimal solution in the search process.Based on the analysis of the principle of CSA,this dissertation improves the traditional CSA and applies the novel CSA to multi-peak function optimization and MIMO blind equalization issues.The main work of this dissertation can be summarized as follows:(1)The concept of multi-peak function optimization problem and the shortcomings of traditional solution are introduced.The CSA,Particle Swarm Optimization(PSO)Algorithm and Genetic Algorithm are carried out through a set of commonly used optimization test functions,respectively.Simulations,based on the comparison and summarization of the experimental data obtained by each algorithm and the actual evolution algebra.It shows that the improved CSA is more suitable for spatial search and solving multi-peak function optimization than the other two algorithms.(2)A direct blind equalization algorithm based agent system is proposed.Using the orthogonal complement projection of the received data and the prior knowledge of the transmitted signal belonging to the finite character set,a quadratic programming problem can be constructed to directly detect the transmitted signal,and then converted into a multi-peak function.The problem is solved by an improved CSA,so that all user signals are recovered,and the improved CSA is compared with the Particle Swarm Algorithm and the traditional blind equalization algorithm.The simulation results show that the improved clonal selection algorithm has the best performance,and can search for the extreme points of the objective function to the maximum extent,maintain diversity,and has strong global search ability.It also verifies the blind equalization ability and good effect of the algorithm.