Self-interference Cancellation In Full-Duplex Massive MIMO Relaying With Precoding

According to the calculation,the total international mobile telecommunication(IMT)spectrum demand in China is 1350?1810 MHz in 2020.At the end of 2014,China has already planned 687 MHz spectrum for IMT.There is at least a 663 MHz deficit.In this context,global research institutions pay great attention to increase the spectral efficiency and then improve the data rate through introducing new technologies for future mobile communication systems.To improve the spectral efficiency,full-duplex systems have been proposed.In theory,the full-duplex mode can double the spectral efficiency compared to the half-duplex mode.The primary challenge in full-duplex transmission is the problem of self-interference.At present,both academia and industry agree that self-interference suppression/cancellation will play the most critical role in implementing full-duplex communication systems.As one of the key alternatives to 5G,massive MIMO technology can exponentially improve the system capacity and make the most of the existing cell sites and spectrum resources.Furthermore,a massive MIMO system has a large surplus of degrees of freedom in spatial domain,which can be used for eliminating effectively all kinds of interferences.Combining with excellent anti-interference properties of massive MIMO technology and taking full advantage of degrees of freedom in spatial domain,this thesis aims at solving the technical bottleneck,self-interference,in full-duplex relaying.In this thesis,the cancellation in full-duplex massive MIMO relaying with precoding is analytically studied over Rician fading channels.An important feature of this method is that no active self-interference cancellation deployed at the relay station,and there is no need of self-interference channel estimation.The main innovative works of the thesis are listed as follows:1)This thesis theoretically demonstrates that in a Rician fading environment,the full-duplex relay self-interference can be perfectly cancelled by zero-forcing(ZF)or maximum-ratio combining/maximum-ratio transmitting(MRC/MRT)processing at the relay station when the antenna number of receive and transmit arrays both tend to infinity.Numerical results indicate that combining with passive self-interference suppression,such as spatial separation,directional isolation,absorptive shielding,and cross-polarization,the self-interference can be effectively cancelled with limited number of antennas and ZF processing at the relay station,even if the self-interference channel has a large Rician K-factor.2)Closed-form expression is approximately derived for the end-to-end achievable rate under the condition of ZF precoding at the relay,which lays a foundation for insightful analyses of the overall system performance.3)The power scaling law of full-duplex massive MIMO systems is investigated.When the antenna number of the transmit and receive arrays at the relay station are sufficiently large,the transmit powers of each source and the relay station can be scaled down proportionally to the number of antennas.Rayleigh channel is a special case of Rician channel,the theoretical analysis results of Rician channels in this thesis hold for Rayleigh channels.Finally,the theoretical analysis is verified by simulation results.