$\mathsf{Hap-SliceR}$: A Radio Resource Slicing Framework for 5G Networks With Haptic Communications

Abstract

It is expected that the emerging 5G networks will not only support diverse use cases, but also enable unprecedented applications such as haptic communications. Therefore, network slicing will provide the required design flexibility. Radio resource slicing would be an indispensable component of any network slicing solution. This paper proposes Hap-SliceR, which is a novel radio resource slicing framework for 5G networks with haptic communications. First, Hap-SliceR derives a network-wide radio resource slicing strategy for 5G networks. The optimal slicing strategy, which is based on a reinforcement learning approach, allocates radio resources to different slices while accounting for the dynamics and utility requirements of different slices. Second, Hap-SliceR provides customization of radio resources for haptic communications over 5G networks. The radio resource allocation requirements of haptic communications have been translated into a unique radio resource allocation problem. A low-complexity heuristic algorithm has been developed for resource allocation. Finally, a comprehensive performance evaluation of Hap-SliceR has been conducted based on a recently proposed 5G air-interface design.