Service providers have now entered the implementation phase for 5G mobile telecommunication networks. With this, the concept of Multi-access Edge Computing (MEC) 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 the platform for malevolent purposes. In this paper we focus on building a realistic 5G-MEC testbed to run legitimate traffic and network attacks, and to collect 5G datasets for 5G-MEC. We also apply a Convolutional Neural Network to the dataset created on our testbed and to publicly available datasets. Our datasets and detection rate suggest that the employment of current public datasets for research based on 5G-MEC security, is now inappropriate.

The exponential growth in the number of devices connected via the internet has led to the need to achieve granular programmability for increased performance, resilience, reduced latency, and jitter. Software Defined Networking (SDN) and Programming Protocol independent Packet Processing (P4) are designed to introduce programmability into the control and data plane of networks, respectively. Despite their individual potential and capabilities, the performance of combining SDN and P4 remains underexplored. This study presents a comprehensive evaluation of SDN with data plane programmability using P4 (SDN+P4) against traditional SDN with Open vSwitch (SDN+OvS), aimed at answering the hypothesis that combining SDN and P4 strengthens the control and data plane programmability and offers improved management and adaptability, which would provide a platform with faster packet processing with reduced jitter, loss, and processing overhead. Mininet was employed to emulate three distinct topologies: multi-path, grid, and transit-stub. Various traffic types were transmitted to assess performance metrics across the three topologies. Our results demonstrate that SDN+P4 outperform SDN+OvS significantly due to parallel processing, flexible parsing, and reduced overhead. The evaluation demonstrates the potential of SDN+P4 to provide a more resilient and stringent service with improved network performance for the future internet and its heterogeneity of applications.

This paper presents a new Intrusion Detection System (IDS) using a 3 layer Convolutional Neural Network (CNN), capable of identifying malicious network traffic. We employ a new injective algorithm to encode network traffic without loss of information. We also include a new algorithm to decode, encoded RGB images back into network traffic. We evaluate the proposed IDS in terms of its computational complexity in for example: time, memory and CPU utilisation for the encoding and decoding algorithms, and its accuracy and loss during training and detection. Lastly, we compare the proposed IDS against a significant IDS algorithm that uses a different approach for encoding, decoding and CNN detection.

As 5G and beyond networks grow in heterogeneity, complexity, and scale, traditional Intrusion Detection Systems (IDS) struggle to maintain accurate and precise detection mechanisms. A promising alternative approach to this problem has involved the use of Deep Learning (DL) techniques; however, DL-based IDS suffer from issues relating to interpretation, performance variability, and high computational overheads. These issues limit their practical deployment in real-world applications. In this study, CiNeT is introduced as a novel DL-based IDS employing Convolutional Neural Networks (CNN) within a bijective encoding–decoding framework between network traffic features (such as IPv6, IPv4, Timestamp, MAC addresses, and network data) and their RGB representations. This transformation facilitates our DL IDS in detecting spatial patterns without sacrificing fidelity. The bijective pipeline enables complete traceability from detection decisions to their corresponding network traffic features, enabling a significant initiative towards solving the ‘black-box’ problem inherent in Deep Learning models, thus facilitating digital forensics. Finally, the DL IDS has been evaluated on three datasets, UNSW NB-15, InSDN, and ToN_IoT, with analysis conducted on accuracy, GPU usage, memory utilisation, training, testing, and validation time. To summarise, this study presents a new CNN-based IDS with an end-to-end pipeline between network traffic data and their RGB representation, which offers high performance and enhanced interpretability through revisable transformation.

Software Defined Networking (SDN) and Programming Protocol-independent Packet Processors (P4) have been attracting attention from both industry and academia as a means of achieving programmability in the network at different layers. The two emerging technologies have in the main, been researched individually in the literature, however, research evaluating the performance of SDN with P4 is limited. As the number of connected users grows and communication continues to evolve, increasing complexity and issues in heterogeneity relating to applications, increase the challenge for service providers to provide resilient connections with faster processing of packets increases. The application of SDN, with its ease of management through centralised programmable control logic, attracts attention from both academia and industry as a means of improving performance. This research has been conducted to explore the hypothesis that combining programmability at both the control plane and data plane, namely SDN+P4, will provide a platform with faster packet processing. In this research, we design a system platform to investigate the network and applications performance in SDN+P4, in comparison to SDN+Open vSwitch. Topologies considered were the multi-path, grid and transit-stub network models with hosts placed at leaf nodes to emulate a small, an intermediate, and a core network. The mininet emulation results demonstrate that across all case studies, parallel processing of P4 has provided more queues for traffic to be processed with the utilisation of a flexible parser. With the evolution of the internet and the heterogeneity of applications, SDN+P4 will, we believe, provide a more resilient and stringent service to all use cases of 5G and beyond.