Machine learning for network intrusion detection on FPGA
Network intrusion detection refers to the monitoring and detection of malicious network activity leading to an attack or unauthorized access to network-connected resources. While an intrusion detection system can identify known attacks, it is a big challenge to detect new or unknown attacks. Furthermore, attackers aim at bypassing intrusion detection systems by adapting their attack strategy to commonly known detection mechanisms. In this project, we will apply machine learning techniques to detect both known and unknown attacks.
FPGAs are increasingly used in commercial network routers and switches to process the received and transmitted data. The benefits of using FPGAs over general-purpose processors are the large amount of input/output pins and the parallel computing capabilities. In this project, we will embed machine learning algorithms on FPGA-based Systems-on-Chip for the purpose of network intrusion detection.