🤖 AI Summary
To address the coupled optimization of base station association and radio resource allocation in ultra-dense networks (UDNs), this paper pioneers the integration of hard backhaul bandwidth constraints into a multi-cell access game model, establishing a non-cooperative game framework that explicitly accounts for backhaul capacity limitations—enabling joint distributed optimization of user association, transmit power, and subchannel allocation. We propose an asynchronous iterative algorithm with rigorously proven convergence guarantees, overcoming resource allocation distortion caused by conventional approaches that neglect backhaul bottlenecks. Under typical dense deployment scenarios, the proposed scheme achieves a 32% increase in system throughput, a 41% reduction in edge-user outage probability, and backhaul constraint satisfaction exceeding 98%, outperforming benchmark schemes. Key contributions include: (i) backhaul-aware coupled modeling of access and resource allocation; (ii) a provably convergent distributed solution mechanism; and (iii) substantial performance gains in both spectral efficiency and reliability.