Wi-Fi-Prioritized Fair Coexistence for 5G/6G Networks Using Markov Modeling

The accelerating reliance on high-speed, ultra-reliable, and low-latency connectivity is reshaping the landscape of next-generation wireless technologies. To meet this increasing demand, 5G and emerging 6G systems increasingly leverage unlicensed spectrum bands already utilized by advanced Wi-Fi 7 deployments. Coexistence between centrally scheduled cellular networks and contention-based Wi-Fi systems in unlicensed spectrum poses significant inter-technology coordination challenges. Differences in MAC (Medium Access Control) protocols between cellular networks and Wi-Fi, particularly disparities in energy detection thresholds, often lead to unfair and aggressive cellular channel access, significantly degrading Wi-Fi performance and increasing harmful interference, especially in densely deployed coexistence environments. This work proposes a novel dynamic spectrum sharing mechanism based on a Multi-Class Wi-Fi-Prioritized Continuous-Time Markov Chain (MC-WP-CTMC) model that optimizes airtime coordination fairness to mitigate Wi-Fi performance degradation. The model prioritizes Wi-Fi access while enabling 5G/6G operators to serve latency-critical and best-effort traffic opportunistically. By incorporating multi-agent deep reinforcement learning within a repeated game framework, it achieves higher overall utility fairness, with a 19% increase in cellular sum utility compared to the baseline Nash equilibrium, while the primary Wi-Fi user attains 9% higher utility. The results demonstrate that the MC-WP-CTMC mechanism achieves Pareto superiority. Also, this study addresses regulatory gaps in the 6 GHz spectrum, where harmonized coexistence frameworks for IMT (5G/6G) and Wi-Fi remain fragmented across regions.