Handling Real-Time Video Traffic in Software-Defined Radio Access Networks

In this paper, we introduce new solutions to handle real-time video traffic, such as video conferencing, in the future software-defined radio access networks. The real-time video is well-known for its high peak-to-mean rate ratio. This is a major challenge for traffic engineering and radio resource allocation, especially in small cell radio networks. We first propose an online method to dynamically estimate the effective rate of video flows, which is the rate the network should support in order to provide a satisfactory quality of experience. Second, traffic engineering methods taking into account characteristics of video flows are presented. Third, a radio coordination method to provide stable video rate across cells is discussed. Fourth, we give a fountain coding scheme to support mobile video users. The proposed solutions are investigated in an ultra-dense small cell network simulator. The simulation results show very significant gains over conventional technologies.