Green small cell operation using belief propagation in wireless networks

In this paper we study the energy efficiency maximization problem focused on finding a suitable set of turned-on small cell access points (APs). Since APs in small cells are randomly deployed and thus redundant in many cases, a mechanism of turning on/off APs is needed. We propose a device-assisted framework that exploits feedback messages from the user equipment (UE). To solve the problem, we apply the optimization method using belief propagation (BP) on a factor graph. Furthermore we propose an online algorithm inspired by BP, called DANCE that requires low computational complexity. Extensive simulations confirm that BP enhances energy efficiency significantly. Furthermore, simple but practical DANCE exhibits close performance to BP, and also better performance than other existing methods and the baseline.