A Distributed Task Rescheduling Method for UAV Swarms Using Local Task Reordering and Deadlock-Free Task Exchange

Distributed task scheduling is an ongoing concern in the field of multi-vehicles, especially recent years; UAV swarm performing complex tasks endows it with new characteristics, such as self-organization, scalability, reconfigurability, etc. This requires to have distributed rescheduling capability dynamically include many unassigned or possible, while satisfying tight time constraints. As one most advanced methods, Performance Impact (PI)-MaxAss algorithm provides important reference for this paper. However, its exchange-based strategy faces deadlock problem, and method should not be limited this. To end, a proposed swarms, which combines local reordering improved exchange strategy. On hand, based on analysis fact that scheduler unreasonable individuals, paper proposes denoted PI-Reorder, simply adds recursive inclusion phase PI-MinAvg algorithm, so can included without relying exchange. other from phenomenon two more vehicles occasionally get caught infinite cycle exchanging same tasks, problem PI-MaxAss analyzed, then solved by introducing deadlock-free strategy, where some defined counters are used detect isolate deadlocks. Then, rescue scenario demonstrate performance PI-Hybrid compared PI-MaxAss. Monte Carlo simulation results show that, PI-MaxAss, only increase number allocations varying degrees, but also reduce average waiting time, ensuring avoidance. The methods secondary optimization existing algorithms escape optima, reconfiguration after adding removing tasks.