Research on Multi-Layer Heterogeneous Task Offloading Strategy in Satellite Assisted MEC System

In certain regions, the limited coverage capability of ground base stations (BSs), as well as their susceptibility to interference or natural disasters, has resulting in weakened hardware capabilities. As a result, they cannot meet the computational task processing requirements of terrestrial mobile users (MUs). In order to meet the low-delay and low-energy consumption requirements of computational tasks for MUs, and while ensuring high throughput, we first introduce a Space-Terrestrial Collaborative offloading architecture based on Low Earth Orbit (LEO) satellites and design an elastic task offloading strategy. Secondly, to capture the designed offloading strategy, the system model composed of MUs, BS, decision satellite (DS) and agent satellites (ASs) is established. Thirdly, by analyzing the established model, performance measures such as delay, energy consumption, forwarding probability, and utilization are derived. Furthermore, a constrained single-objective optimization problem is formulated by balancing delay and energy consumption while considering constraints of stability and load. For facilitating the solution, we transform the problem into an unconstrained optimization problem by applying the penalty function method. Finally, the system strategy parameters are adjusted to evaluate the performance of system in a steady-state, and an ASs-BS-MUs Offloading Optimization (ABMO) Algorithm is proposed to optimize the designed strategy. Experimental results show that the proposed ABMO is more effective in reducing delay and energy consumption.