International Journal of Scientific Research and Engineering Development

Multi-channel blind source separation (BSS) underpins safety- and performance-critical tasks spanning acoustic speaker separation, EEG component extraction, anti-jamming satellite communication, and hyperspectral image unmixing. As channel counts grow and environments become increasingly dynamic, classical independent component analysis (ICA), non-negative matrix factorization (NMF), and deep-learning separators struggle to jointly estimate mixing structure and recover sources with bounded computational cost and graceful adaptation to topology change. Bio-inspired swarm methods have addressed related subproblems: mixing-matrix estimation, sensor-network routing, hyperspectral unmixing, yet a critical gap persists: no existing work systematically connects multi-channel BSS to Physarum polycephalum-inspired network flow optimization, despite a natural correspondence between adaptive-conductance redistribution and the iterative re-weighting BSS requires. This survey bridges that gap. We review the theoretical foundations of Physarum network dynamics and multi-channel BSS, survey recent literature across both domains and their bio-inspired bridges, and propose recasting BSS as a network flow optimization problem over a channelsource graph, amenable to Physarum-style adaptive solvers. Open challenges and a future research agenda are identified for dynamic environments including 5G/6G communications, EEG-based brain-computer interfaces, and remote-sensing image analysis.

Nishant Tripathi, Saniya Shetty, Rudransh Kumar Sah,"Physarum-Inspired Network Flow Optimization for Multi-Channel Blind Source Separation: A Systematic Survey" International Journal of Scientific Research and Engineering Development, V9(3): Page(2405-2414) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.