IoT BOTNET ATTACK DETECTION USING BIG DATA ANALYTICS

Botnets pose a significant and escalating threat to both Internet and network security. The growing ability to identify suspicious online activities has led attackers to adopt more intricate and sophisticated methods of assault. Bots, essentially zombified computers manipulated by a malicious entity, serve as tools for carrying out attacks, spamming, phishing, and extracting information. The primary culprits behind major distributed denial of service (DDoS) botnet attacks have been Internet of Things (IoT) devices for quite some time. This persistent threat persists because numerous manufacturers continue to release IoT products lacking adequate security measures. The vulnerability of IoT devices is exacerbated by their limited memory and computational resources, rendering them susceptible to security breaches. Furthermore, a multitude of security flaws persists due to the insufficient implementation of robust security mechanisms. The existing rule-based detection systems are often inadequate, as attackers find ways to circumvent them. The predominant focus of botnet research revolves around the detection and prevention of malicious bot activities. Effective botnet detection requires advanced analytical capabilities, which are contingent on the nature of the data selected for analysis and the characteristics of the activities being examined. In this study, we aim to employ various algorithms for predicting botnet activity using big data and the Spark framework.