Proactive Mobility Management Based on Virtual Cells in SDN-Enabled Ultra-Dense Networks

Abstract

Ultra-dense networking (UDN) is a promising technology to improve the network capacity in the next-generation mobile communication system. However, it brings in some new challenges to mobility management due to the frequent handovers and heavy signaling overhead. The problem becomes severe for vehicles owing to their fast moving speed, making it more sensitive to the handover delay with reactive handover decision. In this paper, driven by a real-world vehicle mobility dataset, we propose a proactive mobility management solution based on the virtual cell technique for vehicles. Assisted by a trajectory prediction framework based on the long short-term memory neural network, four function modules are designed in the centralized Software-Defined Networking controller to support the proactive solution. The corresponding signaling procedure is then carefully designed, working with virtual cells to reduce the signaling cost. The prediction framework can achieve satisfactory performance of predicting the next location. The proposed proactive solution eliminates the handover delay and reduces the handover signaling cost by 35% compared with the reactive approach.