What question did this study set out to answer?

The aim is to optimize data processing for user devices in remote areas by minimizing delay and energy consumption.

February 22, 2026Open Access

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

Puntos clave

The aim is to optimize data processing for user devices in remote areas by minimizing delay and energy consumption.
Proposed a dual-layer heterogeneous network architecture using UAV and LEO resources.
Modeled optimization as a mixed-integer nonlinear programming problem.
Developed an alternating optimization algorithm to decompose the problem into subproblems: bandwidth allocation, CPU frequency allocation, and task scheduling optimization.
Used convex optimization, Lagrange multiplier method, and ADMM to solve subproblems.
Utilized the Pareto method to identify optimal trade-offs among objectives.
Achieved a 25.8% reduction in total system cost compared to random optimization.
Achieved an 11.63% reduction compared to equal bandwidth allocation.
Achieved a 12.48% reduction compared to the offloading LEO satellite algorithm.
Achieved a 6.84% reduction compared to distributionally robust optimization.

Resumen

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.

Preguntar a la IA

Me gusta

Guardar

Ver artículo completo