ABSTRACT
The fifth-generation (5G) of cellular networks is currently under deployment by network operators, and new 5G end-user devices are about to be commercialized by many manufacturers. This is just a first step in the 5G’s development, and the true potential of 5G is still far from being reached. Currently, one of the main 5G technologies under the interest of the research community is the network slicing. Network slicing will allow exploiting the 5G infrastructure to flexibly and efficiently provide heterogeneous services (e.g., voice communication, video streaming, e-health, vehicular communication). Like every new technology, one of the critical aspects that need to be considered is the security. In this article, we spotlight the security in 5G network slicing. We highlight threats and recommendations, which are presented in terms of life-cycle security, intra-slice security, and inter-slice security. Furthermore, we identify and discuss open security issues related to network slicing.
I. INTRODUCTION
The evolutionary part of 5G, which consists of improvements in the performance compared to 4G, has been already released to the market. But the revolutionary part of 5G is still under the investigation of the research community and in work by the standardization organizations [1], [2]. It aims to provide differentiated services (e.g., voice communication, video streaming, e-health, vehicular communication) by sharing the same 5G infrastructure. This challenging target can be accomplished by using the novel technology of network slicing [3]. Network slicing is one of the topics targeted in the 3rd Generation Partnership Project (3GPP) Release 16 [1], and further addressed in 3GPP Release 17 [2], which are currently in progress with scheduled completion in June 2020 and 2021, respectively. This motivates an investigation of the security aspects of network slicing at this time.
VI. CONCLUSIONS
We have presented threats and recommendations concerning networks slicing security. We conclude that network slicing security brings in a variety of issues that need to be addressed. Because network slicing itself is at an early development stage, the in-depth security analysis is premature. Many open aspects still need to be clarified and further discussed. We refer to some of these and indicate some possible research directions. Among these, we mention end-to-end security, automated defense mechanisms (using artificial intelligence), rigorous implementation and measurement of isolation, and rigorous security models (for network slicing in general or dynamic network slicing in particular). We anticipate still a considerable time until experimental analysis of network slice security can be conducted (at large scale) to validate theoretical results.
