Internet of things forensic data analysis using machine learning to identify roots of data scavenging

Abstract

In this paper, we proposed the blockchain-assisted shared audit framework (BSAF) to analyze digital forensic data in the IoT platform. The proposed framework was designed to detect the source/cause of data scavenging attacks in virtualized resources (VR). The proposed framework implements blockchain technology for access log and control management. Access log information is analyzed for its consistency of adversary event detection using logistic regression (LR) machine learning and cross-validation. An adversary event detected by LR is filtered using cross-validation to retain the precision of data analysis for varying user density and VRs. Experimental results prove the consistency of the proposed method by improving the data analysis, as well as reducing analysis time and the adversary event rate.