LiDAR-Guided Adaptive RSSI Filtering: An Approach for Coordination Target Selection in Multi-Robot Systems

Efficient selective communication is critical in lowcost multi-robot systems to avoid network congestion and high computational overhead. However, identifying collision-risk neighbors quickly and accurately remains a challenge. We propose a LiDAR-guided adaptive RSSI filtering algorithm for coordinating object selection. To stabilize raw RSSI signals, our method adapts the EWMA filter by integrating LiDAR distance information. This ensures signal stability while enabling a rapid response as robots approach. Furthermore, a dynamic RSSI risk threshold is established, triggering communication only when a neighbor's signal strength indicates it is a coordination object. Experiments show the algorithm's responsiveness is significantly superior to baseline methods, including standard EWMA. Its computational time is comparable to EWMA and much lower than complex Kalman filter-based methods, demonstrating enhanced response capability while maintaining low overhead.