Group for Research in Decision Analysis

Dynamic Cloud-RAN topology management - A handovers perspective

Diala Naboulsi Concordia University, Canada

The development of virtualization techniques enables an architectural shift in mobile networks, where resource allocation, or even signal processing, become software functions hosted in a data center. The centralization of computing resources and the dynamic association between baseband processing units (BBUs) and remote antennas (RRHs) provide an increased flexibility to mobile operators, with important reduction in operational costs. With most research efforts focusing on network side metrics, the user experience in such Cloud Radio Access Networks (CRAN) has been neglected. In this work, we make a first step in this direction and we study user handovers in CRAN, showing that a dynamic assignment of computing resources can generate a new type of handovers. However, using a time-varying graph representation, we show that an optimal mapping between BBUs and RRHs can significantly reduce the overall number of handovers. Furthermore, we propose an online BBU-RRH association algorithm, with results similar to an oracle solution requiring perfect knowledge of future traffic demand. We test our algorithms on a real world-based synthetic dataset, showing that we can reduce the number of handovers by more than 20%. Moreover, if a small tolerance for dropped calls is added, more than 30% of handovers can be avoided.


Free entrance.
Welcome to everyone!