5G networks are expected to support various applications with diverse requirements in terms of latency, data rates, and traffic volume. Cloud-RAN(C-RAN) and densely deployed small cells are two of the tools at disposal of mobile network operators to cope with such challenges. In order to mitigate the fronthaul requirements imposed by the C-RAN architecture, several functional splits, each characterized by a different demarcation point between the centralized and the distributed units, have emerged. However, the selection of the appropriate centralization level (i.e., the functional split) still remains a challenging task, since a number of parameters have to be considered in order to make such a decision. In this paper, a virtual network embedding (VNE) algorithm is proposed to flexibly select the appropriate functional split for each small cell. The VNE is formulated as an integer linear programming (ILP) problem whose objective is to jointly minimize the inter-cell interference and the fronthaul bandwidth utilization by dynamically selecting the appropriate functional split. Specifically, dynamic and static ILP-based algorithms are proposed. Finally, dynamic and static VNE heuristics are proposed to address the scalability problem of the ILP-based algorithms in case of dense and ultra-dense mobile networks, respectively.
