Computation Energy Efficiency Maximization for NOMA-Based and Wireless-Powered Mobile Edge Computing With Backscatter Communication

Authors:
Publisher website
Google Scholar
Abstract
In the Internet of Things (IoT) environment, a wide variety of mobile devices (MDs) have become part of it, leading to a dramatic increase in the amount of task data. However, due to the limited battery capacity and computing resources of MDs, a lot of effort is required to be taken on how to process more data with less energy. In this paper, we take into account the low utilization of spectrum resources and the short battery life of the equipment, and a backscatter communication-mobile edge computing (BC-MEC) network system based on Non-orthogonal multiple access (NOMA) communication mode is proposed. In order to maximize the computation energy efficiency (CEE) of the system, we jointly optimize the backscatter coefficient of each MD, the backscatter communication duration, the direct offloading duration, the MEC server processing time, the local processing time, the direct offloading power of each MD, the calculation frequency of the MEC server, and the local calculation frequency of each MD. We then formulate it as a joint fractional optimization problem, which is a non-convex optimization problem that is difficult to solve by heuristic algorithms with high computational complexity. To this end, we transform such a problem into a convex problem and apply the Lagrangian dual method to solve it efficiently. Furthermore, in order to meet different user requirements, two effective iterative D inkelbach algorithms based on B ackscatter C oefficient U pdates (DBCU) are proposed to solve this problem. Extensive simulation results demonstrate the superiority of our proposed approach, which improves the system CEE by at least 10% compared to state-of-the-art methods.
Funding Information
  • National Natural Science Foundation of China (62071327, 62102408)
  • Tianjin Municipal Science and Technology Program (22ZYYYJC00020)
  • Shenzhen Science and Technology Program (RCBS20210609104609044)

This publication has 38 references indexed in Scilit: