Double-Edge-Assisted Computation Offloading and Resource Allocation for Space-Air-Marine Integrated Networks

TL;DR

This work tackles energy-efficient computation offloading in space-air-marine integrated networks by deploying a double-edge architecture where a LEO satellite and UAVs both provide edge computing for MASSs. It introduces a joint optimization framework that optimizes offloading decisions, offloaded workload volumes, and computing-resource allocation at both UAVs and the LEO, solved via alternating optimization and a layered decomposition with MRIS and KKT-based subproblems. The proposed scheme demonstrates significant energy savings under latency constraints and scales to multi-MASS, multi-UAV, and satellite settings, outperforming several benchmarks. The approach offers practical benefits for real-time marine sensing and autonomous operations, with future directions including AI-driven adaptation and more sophisticated mobility models.

Abstract

In this paper, we propose a double-edge-assisted computation offloading and resource allocation scheme tailored for space-air-marine integrated networks (SAMINs). Specifically, we consider a scenario where both unmanned aerial vehicles (UAVs) and a low earth orbit (LEO) satellite are equipped with edge servers, providing computing services for maritime autonomous surface ships (MASSs). Partial computation workloads of MASSs can be offloaded to both UAVs and the LEO satellite, concurrently, for processing via a multi-access approach. To minimize the energy consumption of SAMINs under latency constraints, we formulate an optimization problem and propose energy efficient algorithms to jointly optimize offloading mode, offloading volume, and computing resource allocation of the LEO satellite and the UAVs, respectively. We further exploit an alternating optimization (AO) method and a layered approach to decompose the original problem to attain the optimal solutions. Finally, we conduct simulations to validate the effectiveness and efficiency of the proposed scheme in comparison with benchmark algorithms.

Figures (8)

• Figure 1: Network model.
• Figure 2: Offloading process.
• Figure 3: The proposed solution approach.
• Figure 4: The total energy consumption and overall latency associated with completing $M_{mn}$'s workloads under different values of $a_{mn}$ with fixed $t_{mn}^U=0.4s$ and $t_{mn}^L=0.7s$.
• Figure 5: Illustration of total energy consumption and overall latency under different values of $t_{mn}^U$ and $t_{mn}^L$, respectively.