Detection of P2P Botnets and Creating Smart Grids using Machine Learning and Structured Graph Analysis

Botnets are the bane of modern Information Security. They have been used to carry out major attacks on websites like ESPN, Vogue (Methbot), and e-commerce websites like Alibaba and Amazon. Hence the detection of a Botnet is very crucial and is being heavily researched by cybersecurity experts. This study aims to shed light on a machine learning and graph-based approach for Botnet detection devising our own approach for P2P Botnet discovery. Our Graph-Based Approach includes the identification of suspicious nodes by analyzing their synchronous activities and their inter-nodal connections. An extension of the method proposed in this study is devising a smart grid system. With our Graph and Machine Learning approach, a grid system which supports the bidirectional flow of information and is also self-correcting is proposed.