Estudo, desenvolvimento e implementação de algoritmos de aprendizagem de máquina, em software e hardware, para detecção de intrusão de rede: uma análise de eficiência energética

Abstract

The increasing network speeds, number of attacks, and need for energy efficiency are pushing software-based network security to its limits. A common kind of threat is probing attacks, in which an attacker tries to find vulnerabilities by sending a series of probe packets to a target machine. This work presents the study, development, and implementation of a network packets feature extraction algorithm in hardware and three machine learning classifiers (Decision Tree, Naive Bayes, and k-nearest neighbors), in software and hardware, for the detection of probing attacks. The work also presents detailed results of classification accuracy, throughput, and energy consumption for each implementation.