Load-aware dynamic spectrum access in ultra-dense small cell networks

This article investigates the problem of distributed spectrum resource allocation in the ultra-dense small cell networks, which takes the different loads of small cell base stations (SBSs) into consideration. Here, we simplify the load as the number of active users served by a SBS. We formulate the problem of load-aware channel selection as a graphical game and propose a distributed learning algorithm to achieve stable solutions. With the proposed distributed learning algorithm, SBSs can not only select multiple channels according to their current loads, but also decide their preference to the licensed channels (which are licensed to the macro cells) and unlicensed channels, which contribute to mitigate the cross-tier and co-tier interference. The algorithm is proved to converge to Nash equilibria. Furthermore, the simulation results verify that our proposed learning algorithm can mitigate the cross-tier interference and co-tier interference.