Heuristic task scheduling on heterogeneous UAVs: A combinatorial optimization approach

Abstract

A fleet of unmanned aerial vehicles (UAVs) provide a new and unique type of distributed computing paradigm and platform. This is a distributed computing environment with mobile servers, where a server (i.e., a UAV) moves around to process tasks. Task scheduling for UAVs has several unique characteristics, such as heterogeneity, mobility, and locality. UAVs are heterogeneous in the sense that they have different initial positions, flight speeds, and execution times. While task assignment and flight planning have been studied extensively, there has been little research of task scheduling on heterogeneous UAVs within the framework of combinatorial optimization, i.e., heuristic algorithms for NP-hard problems and their performance evaluation when compared with optimal solutions. In this paper, we take a combinatorial optimization approach to addressing the issues in task scheduling for heterogeneous UAVs. The main contributions are summarized as follows. We define eight combinatorial optimization problems, i.e., four time-centric problems including the completion time minimization problem, the total time minimization problem, the finished tasks maximization with time constraint problem, the reward maximization with time constraint problem; and four distance-centric problems including the longest distance minimization problem, the total distance minimization problem, the finished tasks maximization with distance constraint problem, the reward maximization with distance constraint problem. We prove that all these problems are NP-hard. We develop two efficient and effective heuristic algorithmic frameworks to solve our problems, one for minimization problems and one for maximization problems. We derive lower/upper bounds for the optimal solutions. We evaluate the performance of our heuristic algorithms by comparing their solutions with optimal solutions and show that they are able to produce near-optimal solutions. To the best of the author’s knowledge, this is the first paper which studies task scheduling on heterogeneous UAVs using a combinatorial optimization approach.