Varying Channel Characteristic Parameter Acquisition And Transceiver Design MIMO Systems

This paper studies the acquisition of the time-varying channel characteristicsparameters including fading, frequency offset and time delay for distributed MIMOsystem as well as MIMO relay transceiver design. The contents of the study consist ofthe following issues.For distributed MIMO system, the received signal is a nonlinear function ofthe parameters of the channel characteristics, and it is difficult to estimate theparameters by particle filter algorithms which are based on the nonlinear filteringbecause of the large amount of calculation and the real-time applications. Theefficient importance sampling algorithm was used to solve the sampling issue. Basedon the minimum importance sampling estimation variance criterion, the importancedensity function of high-dimensional numerical integration and bias factor werederived, and then the importance density function were established and thenMetropolis-Hasting algorithm were studied to speed up its convergence. Thesimulation results show that when the initial sampling variance is large, the methodcould obtain preferably mean square error performance.In addition, in MIMO relay system, the joint transceiver could be derived bythe perfect channel state information. Considering the actual channel information isnot perfect, so that we could model the channel estimation error as Gaussianrandom variable of the former channel and the rear channel which is based on therelay in MIMO relay system. Based on the minimum mean square error criterion andthe maximum transmission power of the base station is limited pre-coding matrix,the relay node forwarding matrix and the terminal receives matrix and automaticpower control factor is derived, and iterative algorithm was be used to calculate. Thesimulation results show that this method can lower the mean square error and biterror rate performance at high SNR.