Joint remote radio head selection and user association in cloud radio access networks

Abstract

The cloud radio access network (C-RAN) has been proposed recently as a promising network architecture to meet the explosive data traffic growth in 5G wireless communication systems. In C-RAN, all baseband signal processing is centralized at one Baseband Unit (BBU) pool powered by cloud computing technologies. Whilst the Remote Radio Heads (RRHs), left off on the cell sites, are connected to the BBU pool through fronthaul networks and can be densely deployed with low cost. However, a large number of active RRHs located close to each other may result in severer interference and inefficient energy consumption. To tackle above challenges, we formulate a network power consumption minimization (NPCM) problem, which selects a set of active RRHs and constructs the user association with the active RRHs. The capacity limitation of the fronthaul network is considered in the problem. Since the NPCM problem belongs to the integer programming problem and is NP-hard, we propose a low complexity approximation algorithm that yields the performance guarantees: joint RRH selection and user association (JRSUA) algorithm. A simulation platform is developed to evaluate the network power consumption in three fronthaul network scenarios: fiber, wireless and mixed. Simulation results show that the proposed JRSUA algorithm is able to provide near-optimal performance with reduced complexity and outperforms the other counterparts.