Proceedings of the AAAI Symposium Series
Adaptive Quantum-Inspired Annealing for Scalable Optimization of Wireless Sensor Network Coverage
Amandeep Singh Bhatia, Mandeep Kaur Saggi
Abstract
Wireless sensor network (WSN) coverage is a central optimization challenge in Internet of Things (IoT) deployments, where sensor mobility must be coordinated to maximize target coverage while limiting movement, redundancy, and resource usage. We formulate mobile placement as a quadratic unconstrained binary optimization (QUBO) problem using geometry-derived candidate actions. The formulation is first validated using the quantum approximate optimization algorithm (QAOA) on a 12-logical-qubit, two-sensor, three-target instance. For larger networks, we propose an adaptive simulated quantum annealing-assisted large-neighborhood search method (Adaptive SQA-LNS). The method identifies difficult target regions, solves reduced QUBO subproblems using path-integral simulated quantum annealing (SQA), a classical simulation of quantum annealing, and improves decoded solutions through constraint repair and local search. The experiments study scaling from 12 logical qubits to approximately 471 logical QUBO variables for the 50-sensor, 100-target benchmark. Networks ranging from 5 sensors and 15 targets to 50 sensors and 100 targets are evaluated using a common multi-objective formulation that incorporates coverage, movement, sensor utilization, redundancy, and connectivity. Across the evaluated instances, Adaptive SQA-LNS matches the constraint-programming reference coverage and achieves the smallest objective gap among the SQA-based approaches. These results provide a reproducible framework for studying quantum-annealing-inspired hybrid optimization for scalable wireless sensor network coverage.