The increasing popularity of smart mobile devices has not only aroused an unprecedented growth of wireless data traffic, but also stimulated a fast evolution of mobile applications, which shall inevitably incur a huge amount of electric power consumption and carbon footprint in the 5th generation (5G) wireless networks. Hence, developing green communications and computing systems is of strong needs, and energy harvesting (EH) provides a promising solution. However, powering communication and computing systems purely by renewable energy sources may degrade the quality of service (QoS) due to their intermittent and sporadic nature. Fortunately, hybrid energy supply (HES) where EH and the electric grid coexist, and mobile-edge computing (MEC) emerge as effective remedies for green co...[ Read more ]

The increasing popularity of smart mobile devices has not only aroused an unprecedented growth of wireless data traffic, but also stimulated a fast evolution of mobile applications, which shall inevitably incur a huge amount of electric power consumption and carbon footprint in the 5th generation (5G) wireless networks. Hence, developing green communications and computing systems is of strong needs, and energy harvesting (EH) provides a promising solution. However, powering communication and computing systems purely by renewable energy sources may degrade the quality of service (QoS) due to their intermittent and sporadic nature. Fortunately, hybrid energy supply (HES) where EH and the electric grid coexist, and mobile-edge computing (MEC) emerge as effective remedies for green communications and computing systems, respectively. In this thesis, we will investigate novel design methodologies for HES wireless systems as well as green MEC systems in 5G networks.

First, we investigate the fundamental grid energy consumption and QoS tradeoff in HES wireless networks using a single-user system. A novel performance metric, namely, the total service cost, is proposed to investigate the tradeoff, and base station assignment and power control (BAPC) is adopted to optimize the system. Both optimal and low-complexity sub-optimal algorithms are proposed assuming either non-causal or causal side information. Simulation results shall validate the effectiveness of the proposed algorithms and demonstrate the unique grid energy consumption and QoS tradeoff in HES networks.

Next, we focus on more general HES networks with multiple users. By leveraging the Lyapunov optimization techniques, a low-complexity online BAPC algorithm is developed, which depends simply on the current system state. To determine the network operations, we only need to solve a deterministic optimization problem at each time slot, for which an efficient inner-outer optimization algorithm is proposed. Besides, the proposed online algorithm is shown to be asymptotically optimal both theoretically and numerically.

Using EH receivers is another way to reduce the non-renewable energy demand for wireless systems. To achieve their full potential, we propose a novel ARQ protocol for EH receivers, where the acknowledgement feedback can be adapted based upon the receiver’s energy state. We formulate a throughput-constrained packet drop probability (PDP) minimization problem for wireless links with a non-EH transmitter and an EH receiver, and develop the optimal reception policies. It is shown that the proposed ARQ protocol not only outperforms conventional ARQs in terms of PDP, but can also achieve a higher throughput.

The fourth part of this thesis is devoted to the design methodologies for green MEC systems. In order to provide satisfactory and sustained computation performance as well as achieve green computing, we propose an innovative MEC system with EH-powered mobile devices and develop an effective dynamic computation offloading policy to reduce the execution latency and task failure. It is shown that the proposed algorithm not only achieves asymptotic optimality, but also greatly reduces task failure with minor latency performance degradation. We further consider general multi-user MEC systems with delay-tolerant applications. By adopting the average weighted sum power consumption as the objective, we propose a low-complexity online joint radio and computational resource management algorithm, which has the worst-case performance guarantee and the capability of characterizing the power-delay tradeoff in multi-user MEC systems.