Dynamic Computing Offloading Strategy for Multi-dimensional Resources Based on MEC
Jihong Zhao,Zihao Huang,Xinggang Luo,Gaojie Peng,Zhaoyang Zhu
DOI: https://doi.org/10.1007/978-3-031-20738-9_140
2023-01-01
Abstract:Mobile edge computing (MEC) is a technology that extends network capabilities from the core network to the edge network, which effectively solves the problems of devices in terms of computing capacity, service energy consumption, and service delay. In MEC networks, conventional models cannot meet the actual needs when computing-intensive services arrive dynamically. In addition, the large energy consumption of computing-intensive services makes User Equipment (UE) energy consumption a vital issue. Meanwhile, UE energy consumption is critical to the system economic benefits of the entire MEC network. Aiming at the above problems, a stochastic optimization model that can dynamically combine queue length, computing offloading, resource allocation, and energy consumption control is designed, and a Multi-Dimensional Resource Dynamic computation Offloading algorithm (MDRDO) based on Lyapunov optimization theory is proposed. The algorithm used Lyapunov optimization theory to solve the closed solution of computation offloading, bandwidth resource and computing resource allocation problem, and solved the above stochastic optimization model at the same time. Under the premise of ensuring the stability of the service queue, the algorithm improved the economic benefits of the overall MEC network system. The simulation results indicate that, compared with other algorithms, the proposed MDRDO could save the energy consumption of UE, and at the same time significantly improved the overall economic benefit of the system.