A real‐time attack defense framework for 5G network slicing. (11th February 2020)

• Record Type: Journal Article
• Authors: Bonfim, Michel; Santos, Marcelo; Dias, Kelvin; Fernandes, Stênio
• Is Part Of: Software, practice & experience. Volume 50:Number 7(2020)
• Page Start: 1228
• Page End: 1257
• Publication Date: 2020-02-11
• Subjects: 5G -- bloom filter -- cybersecurity -- network function virtualization -- network slice -- P4; Computer software -- Periodicals; Computer programming -- Periodicals; Computer programs -- Periodicals; 005.3
• DOI: 10.1002/spe.2800
• Languages: English
• ISSNs: 0038-0644

Abstract

Summary: Network Slicing (NS) is a key enabler to support 5G network services on‐demand. However, since NS is a result of the recent advancement in Software‐Defined Networking and Network Function Virtualization, it introduces new security issues which include attacks against an NS instance within an operator network and interslice security threats. In this scenario, identifying and mitigating attacks in real‐time is of paramount importance to improve security aspects. However, it is far from being straightforward. Therefore, this work proposes the FrameRTP4, a P4‐based framework that aims to deliver real‐time attack detection and mitigation mechanisms in 5G NS scenarios. For this, it provides a P4‐based switch that implements an Service Function Chaining protocol layer, an efficient and scalable Access Control List for the detection and mitigation of known attacks, and a monitoring system aiming to reduce the overhead induced on the control channel. Furthermore, it delivers an orchestrator that aims to control all switches in order to enable lifecycle management of NS instances and P4 table rules. Besides, it also performs some autonomous tasks such as the wildcard rules generation and the detection of new threats by using machine learning algorithms. Preliminary results point to the potential benefits of FrameRTP4 to be part of a 5G NS infrastructure.