A Network Intrusion Detection Architecture Based on Class Parallelism on Distributed Switches
Thi-Nga Dao, Huu-Noi Nguyen, Van Son Vu · 2022 13th International Conference on Information and Communication Technology Convergence (ICTC) · 2022
A network intrusion detection model based on a neural network achieves accurate classification performance but suffers from high complexity, which is unsuitable for constrained-resource edge devices. In this work, we adopt classification parallelism and magnitude pruning to construct a lightweight detection model for programmable networking devices. Specifically, the multi-label classification model is decoupled into multiple binary class-specific sub-models, that allows a chain of programmable switches to participate in intrusion detection. Moreover, magnitude pruning is applied to remove weak connections for lowering the model size. Experimental results illustrate that the collaborative detection architecture obtains much lower model complexity than the traditional multi-label classifier without sacrificing in classification performance.