Developing a Testbed with P4 to Generate Datasets for the Analysis of 5G-MEC Security

Service providers have now entered the implementation phase of 5G mobile telecommunication networks. With this realisation.the concept of Multi-access Edge Computing (MC) will play a crucial role when providing services on-the-go with low latency,high availability and high bandwidth. However, due to the low processing power of MEC nodes, adversaries may target theplatform for malevolent tasks. Research at the edge is now taking place in both academia and industry, utilising publiclyavailable datasets in the development of Intrusion Detection Systems and traffic analysis. Research though conducted on a 5G-MEC tested utilising realistic data is still lacking. In this paper we focus on building a realistic 5G-MEC tested to generate apragmatic dataset that can be emploved in future 5G research. The components of our 5G-MEC tested include a homesubscriber server, a mobility management entity, a control and user plan separation gateway, a radio access network, an MECnode, user equipment connected via radio link and a Programming Protocol independent Packet Processing switch. Using ourtested we have run legitimate traffic and network attacks and collected the associated data, generating datasets for 5G-MEC.We have also applied a Convolutional Neural Network to the dataset created on our testbed and to publiclv available datasetsused for detection. Our datasets and detection rate show that the employment of current public datasets for research based on5G-MEC security, is now inappropriate.