Signal-Based Approach for Reliable Drone Swarm Delivery

We propose a failure-resilient Swarm-based Drones-as-a-Service (SDaaS) framework for reliable delivery services. Our framework is specifically tailored to address soft failures, characterized by degradation in drone performance. Failures within drone swarms are an inherent part of real-world scenarios and should be factored to ensure the reliability of services. This framework emphasizes post-assessment of failures as a critical step in preplanning for subsequent service compositions. The framework detects soft service failures from tolerable behavior deviations using margin-aware pruning. Furthermore, a failure severity score is computed using signal-based heuristic. Our aim is to minimize the impact of soft failures on consumers by delivering services within the expected time. This is achieved through formation reordering and service recomposition. Experimental results prove the efficiency of the proposed model with respect to failure correlation scores and service delivery times.