Snake optimizer with oscillating factors to solve edge computing task unloading and scheduling optimization problem
Shi-Hui Zhang,Jie-Sheng Wang,Si-Wen Zhang,Yi-Xuan Li,Yu-Xuan Xing,Yun-Hao Zhang
DOI: https://doi.org/10.1016/j.aej.2024.02.009
IF: 6.626
2024-03-01
Alexandria Engineering Journal
Abstract:The Internet of Things (IoT) has been growing quickly since the 5 G era arrived, which has boosted the quantity of data stream and computational demands. Ultra-dense Edge Computing (UDEC) is a result of the growing popularity of Mobile Edge Computing (MEC) and Ultra-dense Network (UDN), two vital technologies with bright futures. Task offloading is viewed in the UDEC network as a successful strategy to offer mobile users low-latency and flexible computing capabilities. However, the edge cloud's constrained computing capacity and mobile users' dynamically shifting computing requirements present difficulties for the reasonable scheduling of computing requests to the edge cloud. The Power Allocation (PA) problem for mobile users to address this issue. This problem attempts to use quasiconvex technique to minimize energy consumption. Next, the Joint Request Offloading and Resource Scheduling (JRORS) problem is treated as a mixed-integer nonlinear programming problem to minimize request delay and improve welfare. The request offloading problem and the computational resource scheduling challenge are two subproblems of the JRORS problem that can be further broken down, the JRORS issue is therefore viewed as a dual decision-making problem. A Ratio of Power Consumption to Welfare (RCW) problem through mathematical modeling. An improved snake optimizer (SO) is therefore proposed, which was added with oscillation factor to optimize the location of the weight parameter in the location update and the variation of food production. It can enhance the global search capability and local search ability. By conducting simulation experiments, divide the mobile users into nine groups for testing, select energy consumption, welfare, and RCW as the objective functions, respectively, and compare them with other swarm optimization algorithms, such as Differential Evolution (DE), Particle Swarm Optimization (PSO), Ant-Lion Optimizer (ALO), Butterfly Optimization Algorithm (BOA), and Gray Wolf Optimizer (GWO). The improved algorithm achieved minimal energy consumption and optimal welfare in nearly every test case, contributing to the enhanced effectiveness and cost-efficiency of edge systems, thereby achieving the optimization of task offloading and scheduling.
engineering, multidisciplinary