Researchers from South China University of Technology have developed MASTER, a distributed optimization method addressing a critical challenge in sensor networks: matching measurements from multiple sensors to locate multiple targets simultaneously without centralized coordination.
The core problem involves a circular dependency—determining which measurements belong to which targets requires knowing target positions, but localizing targets requires proper measurement association. Traditional approaches either assume matching is pre-solved or demand highly capable individual sensors.
Professor Weineng Chen’s team reformulated localization as a bilevel optimization problem. The innovation separates target position searching from measurement matching, using the deterministic Kuhn–Munkres assignment algorithm for the lower-level association problem while reserving computational resources for upper-level position optimization through particle swarm methods.
The system emphasizes “local search early in the process, whereas cooperation is emphasized later to help the network reach consensus.” Sensors weight contributions based on recent improvements to local solutions, enabling adaptive communication strategies.
Experimental results demonstrated that “MASTER generally produced smaller localization errors than five existing distributed optimization methods,” with advantages increasing as sensor numbers grew. The method proved versatile across multiple measurement types including time-difference-of-arrival and received signal strength data.
Journal: IEEE/CAA Journal of Automatica Sinica
DOI: 10.1109/JAS.2025.125150
Article Title: Multi-Agent Swarm Optimization Method with Contribution-Based Cooperation for Distributed Multi-Target Localization and Data Association
Publication Date: August 3, 2026
Source: Chinese Association of Automation