Signal Processing for Multicell Multiuser MIMO Wireless Communication Systems

Multi-user multi-antenna wireless communication systems have become essential due to the widespread of smart applications and the use of the Internet. Ultra-dense deployment of small cell networks has been recognized as an effective way to meet the exponentially growing mobile data traffic and to accommodate increasingly diversified mobile applications for beyond 5G and future wireless networks. Small cells using low power nodes are meant to be deployed in hot spots, where the number of users varies strongly with time and between adjacent cells. As a result, small cells are expected to have burst-like traffic, which makes the static time division duplex (TDD) frame configuration strategy, where a common TDD pattern is selected for the whole network, not able to meet the users’ requirements and the traffic fluctuations. Dynamic TDD (DTDD) technology which allows the cells to independently adapt their TDD pattern based on the current traffic demand and interference status has been proposed as a solution to satisfy the asymmetric and dynamic traffic demand of small cells. However, one major issue arising from the deployment of small cells with multiple-input multiple output (MIMO) and DTDD technologies is cross-link interference which severely degrades the system’s performance if not properly managed. One encouraging solution is the deployment of reconfigurable intelligent surface (RIS) to flexibly reconfigure the signal propagation and/or the wireless channels between the base stations (BS) and the user equipment (UE). An RIS is a planar surface composed of a large number of low-cost passive reflecting elements, the phase shifts of the RIS elements can be adjusted to meet a certain cost function, e.g., the reflected signals add constructively at the intended users and/ or destructively at the unintended users. In the first part of this thesis we study the integration of an RIS in a multicell multiuser MIMO wireless communication system and propose novel non-iterative algorithms for efficient mitigation of the cross-link interference and maximization of the system-wide spectral efficiency. Furthermore, to reduce the high signaling overhead involved in such a centralized scheme in the collection of all the channel state information (CSI) at the central processing unit (CPU a novel distributed coordinated beamforming algorithm based on the alternating direction method of multipliers (ADMM) technique is proposed. The design objective is to maximize the minimum signal-to-interference-plus-noise ratio (SINR) of the downlink users while satisfying the total power constraint of the downlink BSs and guaranteeing that the maximum interference seen by the uplink users due to the transmission of the downlink cells is below a pre-defined level. Moreover, to further reduce the complexity involved in the joint design of the transmit beamforming vectors and RIS reflection coefficients, we exploit the decoupled nature of the signal-to-leakage-plus-noise ratio (SLNR) maximization problem which allows for a characterization of the solution to the multi-user beamforming problem in terms of generalized eigenvectors. To this end, we propose a low-complexity and non-iterative algorithm for the design of the RIS reflection vector and the Rayleigh-Ritz method for the design of the transmit beamforming vectors. Additionally, we consider the design of robust beamforming vectors that minimize the total transmit power of the downlink cells while satisfying the users’ quality-of-service (QoS) targets in the presence of channel estimation errors. Robust beamforming is considered for the conventional worst-case formulation which has deterministic upper bounds on the norms of the channel estimation errors. We also consider a probabilistic constrained optimization, based on the statistical channel error model. To this end, we adopt an S-procedure and a Bernstein-type inequality to reformulate the problems into tractable ones. 1foreover, we propose two alternating optimization (AO) based algorithms using relaxed semidefinite programming (SDP) to solve these two problems. The evaluations of all the proposed algorithms are carried out using a numerical analysis that shows the effectiveness and performance superiority of the proposed algorithms, as compared to baseline algorithms, in terms of achievable rates, power efficiencies, convergence rate, and complexity. This implies that for the current and future wireless communication systems, the integration of RISs in a DTDD system has the potential to mitigate cross-link interference and the algorithms proposed in this thesis are especially appropriate. In the second part of this thesis we carry out a preliminary investigation on the concept of near-field beamforming. As a result of the combination of massive MIMO and operation at mm Wave frequencies, communicating devices will operate in the nearfield of the BS antenna. We exploit the distance discrimination potential of the nearfield beamforming from a communication perspective to facilitate the deployment of high-rate multi-user MIMO millimeter wave systems. We investigate the system’s performance using several precoding schemes. Our numerical results demonstrate a significant performance improvement due to the capability of the near-field beamforming to support reliable communications even for devices that are located in the same angular direction which corresponds to the “worst case” situation. Furthermore, we study the robust near-field beamforming design. We consider a robust near-field beamforming algorithm that is robust with respect to unknow BS antenna array aperture perturbations. We analytically derive the bounds on the perturbations of the steering vector as a function of the known norm of the coordinate displacements for the ?worst-case robust beamforming scheme. The evaluations of the proposed algorithm are carried out using a numerical analysis that shows the effectiveness and performance superiority of the proposed robust near-field algorithm in comparison with other benchmark schemes. We have proposed and implemented signal processing algorithms for multicell multi-user MIMO wireless communication systems. The algorithms are implemented to address some of the requirements for future wireless cellular networks. These include enhanced spectral and energy efficiencies, reduced latency, and greater flexibility in traffic handling. Our numerical results demonstrate the benefits of our proposed algorithms in addressing these requirements. Therefore, these algorithms are suitable for beyond 5G and future wireless networks. As a result, the work in this thesis is invaluable for network operators, network planners, and the research community.