Distributed Task Allocation for UAV Swarms with Limited Communication

Unmanned aerial vehicle (UAV) swarms have become increasingly indispensable in both military and civilian operations. Task allocation, a crucial aspect of UAV swarm autonomy, involves assigning sequential tasks to each aircraft based on environmental constraints and swarm status. While many task allocation algorithms assume reliable communication among agents, real-world environments often present challenges such as limited bandwidth and message interference. This study presents a new distributed task assignment algorithm for heterogeneous UAV swarms, addressing various task constraints. The proposed auction-based method optimizes total cost, ensures fair workload distribution, and minimizes message size through a two-stage auction process. Comparative evaluations with existing algorithms like CBBA and the central Hungarian algorithm, under the Bernoulli communication model, consider factors such as total task cost, message size, unassignable tasks, and conflict assignments. Results indicate the proposed algorithm's effectiveness in smooth communication environments and its potential advantage in low-bandwidth environments. However, it also highlights potential conflicts in scenarios with communication disruptions. To address deviations due to communication quality, Signal-to-Noise Ratio (SNR) values are monitored throughout task execution.