Large Model‐Driven Network Flow Control for Intellectual Property Talent Management in Collaborative Wireless Communication Systems

ABSTRACT Intellectual property talent management platforms face critical network congestion issues when coordinating distributed intellectual property (IP) professionals, including patent attorneys, trademark specialists, and research analysts, through wireless communication networks. Traditional flow control methods fail to account for diverse IP activity patterns and real‐time collaboration requirements. This study proposes LM‐IPNFC (Large Model‐driven Intellectual Property Network Flow Control), a framework that leverages large models (LM) to classify IP talent activities and optimize network flow control. The framework monitors transmission latency, computational overhead, and queuing delays specific to IP knowledge transfer processes. By combining LM‐based activity pattern recognition with adaptive rate regulation mechanisms, LM‐IPNFC constructs intelligent multi‐node collaborative transmission strategies. Experimental validation against four baseline methods demonstrates that LM‐IPNFC reduces traffic jitter from 0.6–1.2 milliseconds to below 0.2 milliseconds, achieves 125 Mbps throughput representing a 27% improvement over the strongest baseline, and successfully resolves all encountered congestion incidents through context‐aware adaptive responses.