Cooperative Multi-View Sensing with Hybrid Fusion and Channel-Aware Encoding for Environment-Aware 6G Networks

Abstract

Background: 6G deployment is geared toward environment-aware intelligence via Integrated Sensing and Communication (ISAC). Single node sensing is fundamentally limited in terms of occlusion, angular diversity, and interference, thereby proving unreliable in mission-critical tasks. Purpose: This work introduces a Cooperative Multi-View Sensing (CMS) framework that exploits spatially distributed base stations, user equipment, and reconfigurable intelligent surfaces for high-fidelity environment mapping with minimized communication overhead.  Methods: The CMS framework innovatively combines three functionalities: (1) lightweight neural networks at edge nodes are designed to extract semantic features from OFDM waveforms; (2) channel-aware adaptively-encoded transmissions reduce latency by up to 42% at 15 dB by dynamically compressing transmissions according to the present SNR; and (3) a hybrid fusion scheme with distributed edge-level processing combines local sub-maps followed by centralized, multi-view fusion networks for global consistency. The results obtained in an urban intersection and an indoor factory scenario using DeepMIMO ray tracing show its effectiveness. Findings: The CMS achieves an 81.2% latency reduction (293 ms to 55 ms) compared to baseline centralized schemes while surpassing single-node sensing with a 33% accuracy gain. Ablation studies show both adaptive encoding and hybrid fusion are crucial, with performance losses up to 39–116% in latency and 2-10% in accuracy when removed. Scalability analysis indicates linear O(n) scalability, where the latency is 90 ms when 50 nodes are utilized as opposed to centralized (1290 ms) and distributed (622 ms) systems becoming unmanageable. The statistical analysis over 50 Monte Carlo trials confirms robustness (p<0.001) with an optimal cluster size of 8-12 nodes. Conclusion: CMS sets a blueprint for environment-aware 6G wireless communications and facilitates real-time applications with cooperative multi-view sensing. Recommendation: Subsequent 6G deployments must focus on implementing hybrid fusion within the optimal range of 8-12 nodes and employ channel-aware encoding for adaptive compression.