Spectrum Aggregation, Beamforming And Linear Receivers In Wireless Communications:Technologies And Performance Analyses

Discrete spectrum system is consist of a series of noncontiguous idle spectrum segments, spectrum aggregation and precoding are performed by the users, which greatly improve the prac-tical spectral efficiency. Multiple-input Multiple-output(MIMO) is another key technique applied to improve the spectral efficiency, opens up the freedom in space domain besides time and fre-quency, the achieved diversity gain and array gain can greatly increase the system throughput and link reliability. When the number of antennas grows up, Massive MIMO comes into being, com-pared with the traditional MIMO systems, the upper bound of achievable rate can be obtained through linear receivers. Spectrum aggregation, MIMO and Massive MIMO greatly improve the spectrum efficiency and have been the foundation of the Physical Layer of5G wireless commu-nication systems.Similar to the MIMO systems, there exists Diversity and Multiplexing Tradeoff(DMT) prob-lem in discrete spectrum systems, whereas, the method used in analysis of DMT in MIMO sys-tems can be applied. Compared to the low-medium speed scenario, the channel changes quite a lot under high speed, except for the diversity and multiplexing gain, the array gain exists in MI-MO systems, which is applicable in the handover of high speed train communication systems. In Massive MIMO systems, the linear receivers can achieve the upper bound of the achievable rate. In this work, we focus on these three problems in wireless communication system, by adopting the technologies and mathematical analysis method in MIMO systems, we develop some inno-vative strategies and derive some performance analyses, the finite-SNR DMT performance is derived in discrete spectrum systems. Besides, the multi-relay beamforming is adopted in high speed train communication systems to complete the soft handover. Then, the performance of the matched filter(MF) receivers is analyzed in distributed Massive MIMO systems.Firstly, the DMT performance in spectrum aggregated systems is studied. Since after spec-trum sensing process, the idle spectrum appears as a series of fragments, each with different bandwidth and power limit. In order to maximize the ergodic capacity, the optimal power al-location strategy is proposed with total power limit, which is just the inverse power allocation process. Coding can be done independently or jointly of the subchannels, the finite-SNR and infinite-SNR DMTs of these two schemes are studied, respectively. The result shows that when coding jointly among the subchannels, under same multiplexing gain, the diversity gain is greatly improved. Finally, the influence of bandwidth distribution and power limitation to the DMT is studied. Secondly, the multi-antenna-beamforming-based mobile handover strategy and its perfor-mance in high mobility cellular networks are studied. The array gain is achieved in this scheme. High mobility leads to Doppler spread, fast fading and frequent hard handover, by using multi-antenna relay, the complicated channel estimation is no longer needed. Simple but effective statistical optimal beamforming and tapered beamforming are used for communication between the HST and the serving base station(SBS) and target base station(TBS). As the HST moves on, the power allocated to the SBS is gradually decreased while that to the TBS increased, aiming to optimize the overall outage probability or sum rate provided by the two beams. In this way, without explicit connection or disconnection to the BSs, soft handover is achieved. In the calcu-lation of the outage probability and channel capacity, the inter-channel interference (ICI) caused by the serious Doppler effect at high speed is taken into account. Moreover, the beamforming and power allocation parameters are progressively adjusted by exploiting the linear mobility of HST within a small time interval, making the proposed algorithms very efficient.Finally, the performance of MF receiver in distributed massive MIMO systems is studied. When the number of antennas in MIMO system becomes infinite, massive MIMO comes into being and the system performance becomes quite different. The simplest linear receiver, matched filter(MF) receiver, is analyzed in the distributed case here. The result shows that, compared to the co-located case, the noise and pilot contamination part are the same while the multi-user interference is not, it is not only determined by the antenna numbers but also the antenna set spatial distribution. Due to the adjacent antenna set distance requirement and space limit, the number of antenna set cannot be too large, which leads to similar behavior as the co-located case in practical. The achievable rate is dominated by the nearest antenna. Several different typical antenna set spatial distributions are studied finally.