Computation Offloading Strategies for LEO Satellite Edge Computing Systems Based on Different Multiple Access Methods

Abstract

Mobile Edge Computing (MEC) is pivotal for supporting compute-intensive and latency-critical applications in forthcoming mobile networks. It plays an essential role in providing network services through Low Earth Orbit (LEO) satellites, especially in demanding environments. This study proposes an Orthogonal Frequency Division Multiple Access (OFDMA)-based joint optimization framework for Offloading Decision and Dynamic Resource Allocation (ODDRA) Strategy. This framework dynamically allocates computing and bandwidth resources based on the current task load managed by LEO satellites. Moreover, it introduces a Time Division Multiple Access (TDMA)-based joint optimization for Offloading Decision and Task Offloading Sequence (ODTOS) Strategy. This strategy models the task offloading sequence as a permutation flow shop problem, targeting the minimization of total flowtime, and employs the heuristic Liu and Reeves algorithm (LR). The offloading decision challenge is addressed using matching theory and coalition game theory. Simulation outcomes demonstrate that the proposed strategies substantially decrease system delay and energy consumption relative to existing methods.

Keywords

• LEO satellite
• MEC
• offloading decision
• dynamic resource allocation
• task offloading sequence