Secure5G: A Deep Learning Framework Towards a Secure Network Slicing in 5G and Beyond
Anurag Thantharate, Rahul Arun Paropkari, Vijay Walunj, Cory Beard, Poonam Kankariya
Abstract
Network Slicing will play a vital role in enabling a multitude of 5G applications, use cases, and services. Network slicing functions will provide an end-to-end isolation between slices with an ability to customize each slice based on the service demands (bandwidth, coverage, security, latency, reliability, etc.). Maintaining isolation of resources, traffic flow, and network functions between the slices is critical in protecting the network infrastructure system from Distributed Denial of Service (DDoS) attack. The 5G network demands and new feature sets to support ever-growing and complex business requirements have made existing approaches to network security inadequate. In this paper, we have developed a Neural Network based `Secure5G' Network Slicing model to proactively detect and eliminate threats based on incoming connections before they infest the 5G core network. `Secure5G' is a resilient model that quarantines the threats ensuring end-to-end security from device(s) to the core network, and to any of the external networks. Our designed model will enable the network operators to sell network slicing as-a-service to serve diverse services efficiently over a single infrastructure with high security and reliability.