Fuzzy based Mamdani Inference System for Botnet Detection

The word "botnet" blends the words "robot" and "network". The bot master manages the botnet, which allows for remote bot management over the network. The bot master communicates with the bots (infected computers) via IRC C&C channels, from which the bots receive commands and instructions to carry out certain tasks. It spreads to other computers, forming a network. The present paper discusses Mamdani Fuzzy Inference System (FIS) to detect botnets. In contrast to Boolean "true and false" reasoning for fuzzy logic botnet detection, the paper adopts a "degree of truth" approach. The mamdani fuzzy inference system uses the idea of partial truth - where a fact's value might range from totally true to entirely untrue. MATLAB has been used to implement the botnet detection using mamdani fuzzy inference system. The fuzzy regulations were recently adjusted as the data parameters have changed.