Research On Energy Efficiency Optimization Of MIMO Communication Technology

Wireless communication systems focus on not only spectrum efficiency but also energy efficiency due to global energy crisis. Consequently, energy efficiency (EE) is increasingly important to future wireless systems.Multiple Input Multiple Output (MIMO) technology has been widely ap-plied in wireless networks nowadays as a key technology due to the multiplex-ing gain and diversity gain. From EE piont of view, MIMO systems need less transmit power than single-input-single-output (SISO) systems for the same data rate. Therefore, MIMO is proposed as a candidate technology to improve EE of wireless systems. However, previous research on MIMO mainly aims to improve network capacity or spectral efficiency, but rarely concerns EE. Hence, EE in MIMO system has been attracting significant research interests.This dissertation focuses on the EE performance of MIMO system. EE metrics for several application scenarios of MIMO are proposed, which can be used as good basis for future EE works. Base on the proposed EE metrics, we have investigated an energy-efficient channel training scheme in single user MIMO (SU-MIMO) systems, the optimization of EE for BSs with distributed MIMO using an efficient antenna sleep strategy and energy-efficient spectrum sensing strategy in cognitive radio networks. The main contributions of this dissertation are listed as follows:1. An energy-efficient channel training scheme for SU-MIMO systems is proposed.Firstly, we analyse the tradeoff between the accuracy of channel state in-formation (CSI) and total EE. In MIMO systems, imperfect CSI will degrade system capacity. As a result, the performance of EE will be impacted. CSI can be acquired by channel training, which would consume additional energy and time. Therefore, the optimization of channel training under constrains of training power and training time is important for MIMO systems.Secondly, the channel training scheme is formulated as an optimization problem, which jointly optimizes power and time allocation.Thirdly, we analyze the optimization problem and obtain the optimal pow-er and time allocation.Finally, simulation results show that the proposed channel training scheme can effectively improve the total EE.2. A novel metric called sleep EE gain is proposed to evaluate the EE performance of sleep strategy and an efficient antenna sleep strategy for BSs with distributed MIMO is proposed.Firstly, antenna sleep strategy for BSs with distributed MIMO can reduce energy consumption, particularly at low traffic loads, but also cause delay. We studies the approach by which to measure and improve the EE performance of sleep strategy. A metric to evaluate the EE performance of antenna sleep strategies is proposed.Secondly, the distributed MIMO system with sleep mechanism is formu-lated as a multiple vacations queue model with exhaustive service. And then we go through the average power and average request response time of the distributed MIMO systems with sleep mechanism.Thirdly, in order to optimize the proposed metric, we develop a related closed-form expression to determine the optimum sleep period for antennas.Finally, simulation results match the analytical results and show that the proposed strategy can effectively improve the EE in distributed MIMO systems, especially those experiencing low traffic.3. A novel metric is proposed to evaluate the average sensing EE of cogni-tive radio (CR) networks and an optimal cooperative spectrum sensing scheme based on cooperative MIMO is proposed.Firstly, a novel metric is proposed to measure the average sensing energy efficiency performance of the cooperative/coordinated sensing scheme.Secondly, to optimize the proposed metric, we provide a theorem as well as the related closed-form expression for determining the optimum number of cooperative cognitive terminals (CTs). Furthermore, we extend this metric to measure the average sensing EE of the coordinated spectrum sensing scheme. According to the theorem, a CT-assignment strategy for coordinated spectrum sensing is proposed to optimize the overall average sensing EE.Thirdly, an uplink cooperative MIMO scheme is proposed to reduce ener-gy consumption in the reporting stage for optimizing the proposed metric.Finally, simulation results show that the proposed strategy and report scheme can effectively improve the average sensing EE performance of CR systems, especially those containing a large number of CTs.