Abstract:
Cloud Radio Access Network (CloudRAN) is proposed as a candidate architec ture for future mobile networks in order to meet the challenging requirements presented in 5G specifications. The key rationale behind CloudRAN is two fundamental prin ciples; the first one is splitting the base station functionality into two as radio and baseband processing, and the second one is grouping those baseband units (BBU) into centralized and virtualized data centers called BBU pools. CoMP (Coordinated Multi point) is an advanced radio coordination technique taking advantage of using multiple base stations for a single user equipment to improve signal quality and reduce interfer ence. Despite being effectively utilized in current mobile networks, integrating CoMP to the CloudRAN, especially for joint transmission and reception modes, reveals both conveniences and challenges. The most important problem exposes itself as increased pressure over bandwidth requirements due to exchange of raw IQ signals between vir tual BBUs. In this study, the placement of virtual BBUs into physical nodes of BBU pool is explored to discover efficient allocation methods with respect to given CoMP sets. The method for creating an efficient allocation follows the intuition that data ex change between two BBUs causes less stress over the network resources of the BBU pool when they are hosted on the same physical machine, as opposed to exchanging data across different machines. The optimal placement around this insight is formulated and seeing the computational complexity of finding the optimal solution, a heuristic method is developed to produce close to optimal results in polynomial time.
