Nonlinear Parameter Estimation Of Power Amplifer In 60GHz Channel

With the rapid development of digital home appliances,high-definition video services,intelligent terminal and other applications,the demand of transmission rate and signal bandwidth for short distance wireless communication system is stimulated.60GHz millimeter wave communication has been widely used in the future because of its anti-interference and high security,high bandwidth and high transmission rate,and the characteristics of international generality and free license,has become one of the most potential short distance wireless communication systems.In this paper,we analyze and research the problem of parameter estimation of nonlinear power amplifier in 60GHz channel,and achieve some results.To solve the problem of amplifier nonlinear parameter estimation for 60GHz millimeter-wave communication system,researchers have proposed MCMC algorithm and obtain good results.But MCMC methods have certain dependence to initial values,and it searches for optimal value in solution space by one point,so it has slow convergence speed.To conquer the shortage of MCMC algorithm,Genetic Algorithm and Artificial Fish Swarm Algorithm are applied to nonlinear parameter estimation in 60 GHZ communication system for the first time.With the help of strong global search ability and faster convergence speed of Genetic Algorithm and Artificial Fish Swarm Algorithm,nonlinear parameter estimation and detection problem can be solved effectively.The error of the estimated results and the actual results obtained through simulation experiments are within reasonable limits.Thus GA and AFSA can be applied in 60GHz wireless communication system to estimate nonlinear parameter and detect signal with estimated result.The main innovation of this paper is that,for the first time,GA and AFSA are introduced into nonlinear parameter estimation in 60GHz wireless communication system.The features of this two algorithm are analyzed and compared at the same time.The results of the analysis shows that the single run process of GA is fast,but its convergence speed is slow.However,the process of single run for AFSA is more complicated than GA,but its convergence speed is faster than GA,so it can obtain estimated parameters rapidly.