Towards intelligent virtual resource allocation in UAVs-assisted 5G networks

摘要

With the aim of providing novel services and applications having high data rate and low latency, the 5G networks are designed. Thus, it is essential to manage and schedule the physical resources efficiently in 5G era. Virtualization technologies are seen as the most potential enabling approaches towards 5G networks. The known key issue is the virtual resource allocation, called as virtual network embedding (VNE). However, previous researchers simply focus on allocating resources in an inflexible way. As unmanned aerial vehicles (UAVs) communication plays an important role in 5G, we need to add UAVs into the 5G networks in order to expand the coverage and agility of services. In this paper, we investigate the issue of intelligent virtual resource allocation in UAVs-assisted 5G networks. Especially, we propose one intelligent virtual resource allocation algorithm, labeled as Intell-UAV-5G. When receiving one network service, our Intell-UAV-5G implements the network service in an intelligent manner. Though virtualized UAV moves, our Intell-UAV-5G is able to predict all possible connecting access nodes and continue the network service intelligently. Meanwhile, our Intell-UAV-5G can minimize the time of service interruption. Thus guaranteeing the service quality. We also conduct the experiment to highlight the Intell-UAV-5G efficiency. The mostly-similar algorithms are derived for performance comparison. Performance metrics are carefully recorded and discussed.