Efficient User Association in Cellular Networks With Hybrid Cognitive Radio Relays

Hybrid cognitive radio (CR) relays serve cellular users in a two-hop fashion to improve the network capacity-the communication with base station (BS) operates in cellular spectrum, while that with users operates in unlicensed spectrum for dynamic spectrum access. In this letter, we present a user association framework that optimizes the overall system downlink rate in cellular networks using hybrid CR relays, with an emphasis on dynamic load balancing among the BS and the CR relays given dynamically changing available resources. Users select their preferred relays on the basis of their link and resource condition observations, and meanwhile, relays perform admission control based on the estimated number of users associated with each relay at Nash-stable state, which is an indicator of the expected load condition of the overall network. Simulation results demonstrate that the proposed scheme outperforms the existing schemes while incurring a small computational complexity.