Research On Interference Alignment With Imperfect CSI

In recent years,people's demand for higher data rates has promoted a rapid development of mobile communication technology.With the improvement of the transmission rate,limited spectrum resources become increasingly scarce.Multiple antennas technology is a dramatic technology that can greatly improve the spectral efficiency and transmission rate.However,multiple antennas result serious interference,which makes the system an interference-limited system.Interference alignment provides a new way for researchers to manage the interference,which can convent the interference to one subspace,so that there will be enough space for effective signal.Therefore,interference alignment has been widely studied a s soon as it was put forward.However,in realistic communication scenarios,it's difficult to get perfect CSI,which will affect the performance of the existing interference alignment method seriously.Firstly,this paper introduces several commonly used channel models of multiple user MIMO system,and studies some basic theory and the implementation of IA.Then the validity of IA is verified by analyzing the simulations of three classical IA algorithms--Max-SINR,Min-Leakage,MMSE.And then in order to overcome the disadvantage of some c lassical algorithm,a combined IA algorithm with low complexity is proposed,the analysis and the simulation results show that the proposed algorithm achieves a better performance in terms of complexity.Secondly,this thesis analyzes the performance of the IA algorithms in both perfect and imperfect CSI,especially the impact of the CSI with estimation error or delayed feedback on the system.Two IA algorithms in imperfect CSI are investigated,including Max-SINR and Min-Leakage.And then an improved MMSE algorithm of interference channel is proposed to deal with the estimation error,the simulation shows that the performance varies with the estimation of the CSI.Finally,a MMSE algorithm is introduced in ideal MIMO X channel.However,the performance of this MMSE algorithm is poor in realistic communication scenarios without considering the estimation error.So this paper proposes a novel MMSE algorithm,which can deal with the estimation error properly.With this proposed algorithm,the system could manage the interference better in imperfect MIMO X channel.