Investigation Of Interference Suppression And MIMO Precoding Techniques For The Cellular Wireless Communication Systems

To better support the emerging enhanced mobile broadband services,such as high definition video and virtual reality,the performance of the fifth generation(5G)cellular wireless communication systems will be further improved in terms of peak data rate,user experienced data rate,spectral efficiency(SE),energy efficiency(EE)and so on.To achieve these goals,cell densification,millimeter wave(mmWave),massive multiple input multiple output(MIMO),and full duplex(FD)have been regarded as the key technologies for 5G.However,the inter-cell interference is more severe in dense networks.The power of the self-interference(SI)is very strong in FD systems.Moreover,mmWave can easily be integrated with massive MIMO for improved link reliability,but the multiuser interference should be handled with the constraints of low hardware complexity and energy consumption.Therefore,the interference in 5G systems can not be effectively suppressed by the conventional inter-cell interference management techniques.The interference has become one severe issue that limits the system performance,due to the fact that it has more diverse types,stronger power,more random topology,and more constrained suppression ability imposed by the low hardware complexity.In this thesis,some novel interference suppression schemes and MIMO precoding algorithms are investigated for the multi-cell systems,FD systems,and the mmWave massive MIMO systems,respectively.Firstly,the interference suppression and MIMO precoding algorithms for the downlink cooperative multicell systems are investigated.In contrast to the classical interference alignment(IA)algorithm with minimizing the interference power,a novel IA algorithm is proposed to maximize the received signal-to-interference-plus-noise ratio(SINR).It is proved that the precoding and receive matrices can be designed based on the generalized eigenvalue decomposition of the received signal covariance matrix and the interference plus noise covariance matrix.Therefore,it strikes a balance between the maximization of the signal power and the minimization of the interference and noise power.Simulation results show that the sum rate of the systems is increased considerably at the low-to-intermediate signal-to-noise ratio(SNR)region.In addition,an interference decoding scheme for the uplink cooperative multicell systems is proposed,based on the idea of sparse code multiple access(SCMA)technique.The cell-edge uplink users in different cells can share the same resources,however,are distinguished by different children codebooks.The base stations(BSs)decode both the useful signal and the interference simultaneously.Simulation results show that the proposed schemes can achieve better bit error rate(BER)performance than the fraction frequency reuse scheme in 4G systems.Since the EE becomes a main concern of the 5G systems,the criterion of maximizing the minimum EE of the downlink cooperative multicell systems is considered.It is shown that the precoding design can improve the EE of the systems at high SNR and guarantee the EE fairness among different BSs.Secondly,the hybrid precoding design for downlink multiuser mmWave massive MIMO systems is investigated.When the systems operate in the frequency division duplex mode and the phase-shifting network has fully-connected structure,the minimum mean square error criterion is considered.When the number of antennas at the BS goes large,it is proved that the optimal analog processing can be designed by the dominant direction of arrival(DoA)of the users during the uplink transmission,therefore,reducing the feedback overhead from the downlink users.In addition,when the angle mismatch between the dominant direction of departure(DoD)in the downlink and the dominant DoA in the uplink has a truncated Laplacian distribution,the closed-form approximation expression for the average signal power loss is derived.When the systems operate in the time division duplex mode and the phase-shifting network has partially-connected structure,a low-complexity iterative hybrid precoding algorithm is proposed with proved convergence to maximize the weighted sum rate of the systems.Besides,its application is extended to the cases of fine-resolution phase shifting networks.Simulation results show that under the restrictions of low hardware complexity and power consumption,the proposed hybrid precoding algorithms can suppress the multiuser interference effectively and reduce the sum rate loss of the systems.Thirdly,the MIMO precoding design for the heterogenous networks is investigated to suppress the inter-layer interference and the SI of the FD backhauling links.The small cells use FD technique to exchange the backhauling information,and reuse the downlink spectrum of the half-duplex macro BS to increase the SE of the access networks.It is assumed that the small cells are equipped with a particular FD precoding structure,in which the auxiliary radio frequency chains are added to increase the spatial degree of freedom and enhance the ability of interference suppression.The MIMO precoding matrices at the macro BS and the small cells are designed to maximize the weighted sum rate of the heterogenous networks,where the weight is used to adjust the priority of the access link and the backhauling link.A distributed iterative algorithm is proposed with proved convergence.Particularly,limited information exchanged is required between different BSs during the iteration process of the proposed algorithm,therefore,reducing the signaling overhead.Simulation results show that compared to the half-duplex mode,the FD technique can improve the capacity of the backhauling link.In addition,the performance gain is determined by the power of SI,the cancellation power of the auxiliary paths and the number of transmit antennas at the FD small cells.Finally,the average achievable rate and the power allocation schemes of FD small cell systems with massive MIMO is investigated with the assumption that the SI channel and the multiuser channel are Rayleigh distributed and Rician distributed,respectively.The approximate closed-form expressions of the achievable rate for both the zero-forcing and maximumratio transmission/maximum-ratio combining processing at the BS are derived.Moreover,the effect of the number of antennas,transmit power,the hardware impairment and the Rician factor on the power of the SI are analyzed.It can be shown that the power of SI can be reducedsignificantly as the number of transmit and receive antennas at the BS increases.In addition,an iterative power allocation algorithm is proposed with proved local convergence to optimize the transmit power of the BS and the uplink users.The effect of the SI and the multiuser interference on the sum rate of the systems is reduced and the SE of the systems is improved significantly at high SNR region by the proposed power allocation algorithm.