Residual Energy-Based Computation Efficiency Maximization in Dense Edge Computing Systems

With the rapid development of 5G, artificial intelligence, the internet of things (IoT) and other technologies, the number of intelligent terminal devices is growing explosively, bringing huge challenges to the existing communication network and cloud computing service mode. The dense edge computing system (DECS), which combines mobile edge computing (MEC) with an ultra-dense network (UDN), has the potential to significantly improve low latency of communications and enhance the quality of experience (QoE) of user equipments (UEs). In this paper, to achieve energy-efficient MEC, computation efficiency (CE) is maximized by jointly optimizing computation offloading, subchannel allocation and power allocation, which yields a challenging non-convex problem. Specially, due to the heterogeneity of UE battery capacities and residual energy, the residual energy of UEs should be taken into consideration in order to achieve better QoE. Therefore, we develop a residual energy-based computation efficiency (RECE) optimization scheme to maximize CE, where the optimization problem is divided into three subproblems. Firstly, the computation offloading subproblem is addressed by a many-to-one matching strategy. Secondly, the subchannel allocation subproblem is dealt with by adopting the graph coloring algorithm. Finally, the power allocation subproblem is solved by the concave-convex procedure (CCCP) method. The numerical results illustrate that UEs' CE can be optimized based on their residual energy in the proposed RECE scheme. Additionally, compared to a scheme without considering UE residual energy, the system CE can be much enhanced, and the UE energy consumption can be significantly reduced in the RECE scheme.