Channel-Adaptive Resource Allocation Architecture for Resource-Efficient RAN Slicing

Serae Kim, Sunghyun Jin, Junseon Kim, Kyunghan Lee

Abstract: Network slicing is one of the key techniques for realizing future services with diverse requirements. In radio access network(RAN), dynamic slice management for guaranteeing quality-of-service(QoS) of each user is especially challenging due to highly fluctuated radio channel for each user. To cope with such channel changes, it is inevitable to allocate more radio resources than are actually used(i.e., over-provisioning). To handle this resource inefficiency, there have been RAN slicing studies for adaptively allocating resources according to channel changes. However, their approaches, which focus on ensuring average performance over time and users, make it difficult to provide guaranteed performance for each user. In this paper, we propose channel-adaptive resource allocation architecture that enables resource-efficient per-user performance guarantees. The proposed architecture consists of two main components: 1) inter-slice allocation which tries to secure resources to satisfy each slice's service-level agreements and 2) intra-slice allocation for per-user resource allocation for satisfying performance requirements of each service. This paper concludes with a discussion on open issues to practical usage of the proposed architecture.