Particle Swarm Optimization Algorithm And The Application In Blind Equalization

Particle swarm optimization algorithm is one of swarm intelligent algorithms that imitate the foraging and flying behavior of birds. The concept of this algorithm is simple. It has few parameters and can be easily achieved. It is an efficient search algorithm which is now widely used in system design, multi-objective optimization, pattern recognition, signal processing, decision making, robotics applications, etc. In the communication systems which can not get long enough training sequences, blind equalization and blind detection can improve system performance greatly. Blind equalization algorithm has become research focus in communication signal processing in recent years. This paper mainly studies the basic and improved particle swarm optimization algorithm and their application in the blind equalization.This paper is organized as follows. The first chapter generally introduces the background, the development of blind equalization technology and the main work of this paper. The second chapter describes the theory, application and research directions of particle swarm optimization, emphasizing on the inertia weight in the algorithm. The third chapter illustrates the theory of blind equalization and MIMO technology. The fourth chapter introduces and simulates blind detection algorithm based on the basic Particle Swarm Optimization. The fifth chapter is the main point of this paper. In this chapter, blind detection algorithm based on simulated annealing particle swarm optimization algorithm and chaotic particle swarm optimization is constructed, and the influence of annealing factor and different chaotic mapping is analyzed. By comparison, a modified algorithm - based on simulated annealing chaotic particle swarm optimization algorithm for blind detection is proposed. Moreover, in order to ensure good performance and low complexity, another improved algorithm called blind detection algorithm based on self-regulation particle swarm optimization is presented. Finally, the BER performance of the proposed algorithm is simulated, and the complexity of this algorithm is analyzed. Simulation shows that the proposed two algorithms, compared to the basic algorithm, has low error rate, fast convergence and stable performance, and can be a good solution to the blind detection problems.