Joint Task Offloading and Resource Allocation Optimisation for BS-Impaired UAV-Assisted Edge Computing
Aofei Dong, Yaowen Hu, Zhixin Mei, Kuiyuan Feng · 2024
With the continuous advancement of wireless communication technologies, the negative impact of base station (BS) impairments on network performance is becoming more and more significant. In this paper, we explore how task offloading and resource allocation optimisation can be achieved through Unmanned Aerial Vehicle (UAV)-assisted edge computing in the context of BS impairments. This study proposes a UAV-assisted vehicular edge computing architecture that aims to provide the required Quality of Service (QoS) for computationally intensive and delay-sensitive applications. We provide an in-depth study of the task offloading and resource allocation problem with the aim of maximising the benefit gained by the vehicle by offloading tasks. This benefit is quantified by measuring the weighted sum of task completion time and energy consumption. We are faced with a mixed integer nonlinear programming (MINLP) problem that involves the joint optimisation of the task offloading decision, the uplink transmission power of the mobile vehicle, and the computational resource allocation on the UAV. Given the complexity of the problem, finding an optimal solution is both difficult and unrealistic for large-scale networks. To address this challenge, we employ a decomposition strategy that splits the original problem into two subproblems: a resource allocation (RA) problem that fixes the task offloading decision, and a task offloading (TO) problem that optimises the optimal value function corresponding to the RA problem. We solve the RA problem using convex optimisation and proposed convex optimisation techniques and apply genetic algorithms to solve the TO problem. The results of simulation experiments show that our proposed algorithm is close to the optimal solution in terms of performance and significantly improves the vehicle unloading benefits when compared to the conventional methods. This suggests that UAV-assisted edge computing can effectively optimise task offloading and resource allocation even when the base station is compromised, thus improving the overall performance and quality of service of the network.