Energy-Efficient Resource Allocation in Multi-UAV-Assisted Two-Stage Edge Computing for Beyond 5G Networks

TL;DR

This paper proposes an energy-efficient resource allocation framework for a multi-UAV-assisted two-stage edge computing system in beyond 5G networks, optimizing task offloading and resource management to reduce energy consumption.

Contribution

It introduces a novel joint optimization approach for task offloading and resource allocation in UAV-assisted MEC systems, employing BSUM to handle non-convexity and improve energy efficiency.

Findings

Proposed framework reduces energy consumption of mobile devices and UAVs.

Achieves near-optimal solutions with guaranteed convergence using BSUM.

Demonstrates superior performance over baseline methods in extensive evaluations.

Abstract

Unmanned aerial vehicle (UAV)-assisted multi-access edge computing (MEC) has become one promising solution for energy-constrained devices to meet the computation demand and the stringent delay requirement. In this work, we investigate a multiple UAVs-assisted two-stage MEC system in which the computation-intensive and delay-sensitive tasks of mobile devices (MDs) are cooperatively executed on both MEC-enabled UAVs and terrestrial base station (TBS) attached with the MEC server. Specifically, UAVs provide the computing and relaying services to the mobile devices. In this regard, we formulate a joint task offloading, communication and computation resource allocation problem to minimize the energy consumption of MDs and UAVs by considering the limited communication resources for the uplink transmission, the computation resources of UAVs and the tolerable latency of the tasks. The formulated problem is a mixed-integer non-convex problem which is NP hard. Thus, we relax the channel assignment variable from the binary to continuous values. However, the problem is still non-convex due to the coupling among the variables. To solve the formulated optimization problem, we apply the Block Successive Upper-bound Minimization (BSUM) method which guarantees to obtain the stationary points of the non-convex objective function. In essence, the non-convex objective function is decomposed into multiple subproblems which are then solved in a block-by-block manner. Finally, the extensive evaluation results are conducted to show the superior performance of our proposed framework.