TRMaxAlloc: Maximum task allocation using reassignment algorithm in multi-UAV system

Abstract

The on-demand flexibility of UAVs and their line-of-sight communication capabilities have made them an effective solution to provide emergency services such as search and rescue in emergencies. The maximum task allocation to a multi-UAV system under time constraints has gained significant interest from academia and industry to enhance the quality of service requirements. To accomplish this goal, we propose the maximum task allocation (TRMaxAlloc) algorithm, which works in two phases: task assignment and reassignment. The first phase assigns the task using the performance impact (PI) algorithm. In the second phase, the assigned task is reassigned to other UAVs to create feasible time slots for the unassigned tasks, increasing the number of assigned tasks. The reassignment of tasks is based on a novel task costing mechanism, which is the main difference between the two assignment methods. The paper identified task selection during the reassignment phase as a critical parameter for maximizing tasks and proposed three methods for task selection. Following that, to evaluate the performance of the proposed TRMaxAlloc algorithm, numerical results are compared with other benchmark schemes under different task selection techniques. The simulation results confirm an improvement of up to 22.41% in task allocation compared to the PI algorithm and 2.41% compared to the PI-MaxAss algorithm under various constraints.