Joint task offloading and resource allocation for remote IoT in space-air-ground networks architecture

In remote areas with insufficient ground infrastructure, user devices (UDs) are constrained by limited computing resources, which poses substantial challenges to achieving low-latency and energy-efficient data processing. To address these issues, this paper proposes a dual-layer heterogeneous network architecture that makes full use of unmanned aerial vehicle (UAV) and low Earth orbit (LEO) computing resources. Considering the high mobility of LEO satellites, the characteristics of channel variations, and the queuing delay of user task offloading, the optimization objective is modeled as a mixed-integer nonlinear programming problem, aiming to minimize the weighted sum of delay and energy consumption (i.e., the total system cost). A low-complexity alternating optimization algorithm is proposed. The original problem is decomposed into three subproblems: bandwidth allocation, central processing unit (CPU) frequency allocation, and task scheduling optimization, which are solved using convex optimization, the Lagrange multiplier method, and the alternating direction method of multipliers (ADMM), respectively. Finally, the Pareto method is used to seek the best trade-off among the optimization objectives. The simulation results indicate that the average total system cost of the alternating optimization for task offloading and resource allocation (AOTORA) decreases by 25.8%, 11.63%, 12.48%, and 6.84% compared with random optimization, equal bandwidth allocation, the offloading LEO satellite algorithm, and distributionally robust optimization, respectively.