Particle Filtering Based Estimation For MIMO Time-varying Channel

With the development of mobile communication,the demands for the limited frequency resource are booming.The major challenge in today's wireless communication is how to serve the explosively increasing demand of multimedia service within the limited bandwidth.Among those emerging technologies, multiple-input multiple-output(MIMO) communication architecture has showed its infinite potential to improve spectral efficiency dramatically.By deploying multiple antenna elements at both transmitter and receiver,MIMO has been regarded as an inevitable promising approach to exploit maximum spatial resources among many emerging technologies.It can enormously increase the wireless communication system capacity and the data rate without an increase in bandwidth and transmission power, thereby providing much higher bandwidth efficiency than the traditional single-antenna system.To achieve a MIMO system,the receiver not only need the low-complexity and high-performance detection algorithm,but also need the channel estimation of MIMO. MIMO channel has far more parameters than the traditional single-input single-output (SISO) channel,so channel estimation is very complexity,and it is difficult to achieve for the receiver.Monte Carlo algorithm based on recursive propagation.A particle set, which is randomly sampled from probability function and has corresponding weights,is introduced to approach the posterior distribution.Therefore it can handle nonlinear and non-Gaussian problems without any limits.It is the main method to deal with non-linear non-Gaussian problem.Based on the outcome of previous study,the thesis focuses on particle filter algorithm and applied it to multiple-input multiple-output channel estimation.In this thesis,at first,based on the wireless channel,it has mainly made research on the related theories about three main aspects:channel models,channel capacity, channel estimation and implementation algorithms.And it analyzed the characteristics of MIMO channel,established the mathematical model of MIMO channel,introduced the traditional channel estimation algorithm of MIMO.Based on the theory of Bayesian estimation and Monte Carlo methods,we introduce the particle filtering.Through establish the state-space of MIMO channel estimation,particle filter is used in channel estimation of MIMO time-varying system,that is the particle filter channel estimation algorithm.Finally,simulated to compare the particle filter channel estimation algorithm with the traditional method of channel estimation performance.The simulation results show that,compare with the traditional channel estimation methods,the particle filter can reduce the computation,at the same time,it can reduce the error rate and minimum mean square error,and proved the advantage of particle filter algorithm.