A Distributed Task Bundling Allocation Algorithm for Multi-UAV under Non-Ideal Communication Conditions

In recent years, multi-unmanned aerial vehicle (multi-UAV) technology has achieved significant progress, demonstrating extensive applications in collaborative reconnaissance and surveillance, search and rescue, and logistics operations. Distributed task allocation, a key component of multi-UAV coordination, enhances task effectiveness through inter-UAV situational awareness exchange while reducing reliance on centralized computing nodes. However, practical implementations often experience non-ideal communication conditions such as packet loss and data errors, which degrade the performance of distributed task allocation methods. To address this challenges, particularly high communication overhead, large convergence runtime, and high conflict probability caused by packet loss and data errors, we propose a novel static distributed task bundling allocation algorithm considering communication compensation (DTBA-C3), and provide convergence proof. Developed from the Consensus-Based Bundle Algorithm (CBBA) framework, DTBA-C3 introduces two key components: 1) an error-filtering strategy with message reorganization to minimize conflicts from data errors, and 2) a packet loss compensation strategy employing maximum bid consensus to estimate lost information and reduce retransmission needs. Furthermore, we conducted numerical simulations to evaluate the algorithm’s conflict probability, communication overhead, onvergence runtime, and objective function value against four benchmarks: CBBA, Hybrid Information Plan Consumption algorithm (HIPC), Performance Impact algorithm(PI), and Decentralized Genetic Algorithm (DGA). The simulation results demonstrate that under non-ideal communication conditions with data errors and packet loss, DTBA-C3 achieves a 22.5% reduction in communication overhead and approximately 50% faster convergence compared to baseline methods, while maintaining conflict-free allocation capabilities and ensuring allocation optimality.