Distributed Multi-Objective Dynamic Offloading Scheduling for Air-Ground Cooperative MEC

TL;DR

This work tackles sequential, multi-objective offloading and trajectory planning in air–ground cooperative MEC with a UAV-mounted edge server. It introduces a distributed MORL framework that uses kernel-based function approximation and an $n$-step return to jointly optimize UAV trajectory and per-UE offloading decisions, addressing the curse of dimensionality from multiple UEs. Results show the kernel-based approach with $n$-step returns achieves lower long-term backlog and energy consumption and substantially faster decision-making and online learning than a fully-connected DNN baseline, while effectively leveraging LoS air–ground channels. The proposed method enables online learning with feature expansion, enhancing performance in highly dynamic MEC environments.

Abstract

Utilizing unmanned aerial vehicles (UAVs) with edge server to assist terrestrial mobile edge computing (MEC) has attracted tremendous attention. Nevertheless, state-of-the-art schemes based on deterministic optimizations or single-objective reinforcement learning (RL) cannot reduce the backlog of task bits and simultaneously improve energy efficiency in highly dynamic network environments, where the design problem amounts to a sequential decision-making problem. In order to address the aforementioned problems, as well as the curses of dimensionality introduced by the growing number of terrestrial terrestrial users, this paper proposes a distributed multi-objective (MO) dynamic trajectory planning and offloading scheduling scheme, integrated with MORL and the kernel method. The design of n-step return is also applied to average fluctuations in the backlog. Numerical results reveal that the n-step return can benefit the proposed kernel-based approach, achieving significant improvement in the long-term average backlog performance, compared to the conventional 1-step return design. Due to such design and the kernel-based neural network, to which decision-making features can be continuously added, the kernel-based approach can outperform the approach based on fully-connected deep neural network, yielding improvement in energy consumption and the backlog performance, as well as a significant reduction in decision-making and online learning time.

Figures (5)

• Figure 1: Air-ground collaborative MEC with a UAV and a BS.
• Figure 2: Number of task bits produced in each UE over timeslots.
• Figure 3: Performance achieved by kernel-based approaches with $n$-step return (for $n=30$) at different weights and 1-step return. (a) Average energy consumption as a function of timeslots. (b) Average backlog of task bits as a function of timeslots.
• Figure 4: Performance comparison of the kernel-based approach and the DNN-based approach after running for 7000 timeslots. (a) Instantaneous energy consumption as a function of timeslots. (b) Instantaneous backlog of task bits as a function of timeslots.
• Figure 5: The UAV's trajectory, where the notations $(x, y)$ and $t$ represent the horizontal position of the UAV and the index of a timeslot, respectively.