Preventing from Cross-VM Side-Channel Attack Using New Replacement Method

Abstract

As Cloud services are gaining importance, many recent works have discovered vulnerabilities unique to such systems. Specifically, like it promotes a risk of information leakage across virtual machine isolation via side-channels. Cloud environment allows mutually distrusting clients access to the shared hardware simultaneously, which can be termed as the main reason for a side-channel attack (SCA). This paper tries to investigate the current state of side-channel vulnerabilities involving the central processing unit cache and identifies the shortcomings of earlier defenses in a Cloud environment. Through cache-based SCA, fined grained information can be collected by attacker easily, and this information may be used by the attacker to infer meaningful results like a secret key, etc. In this article, we detect the SCA at an earlier stage through flush-reload based statistical techniques which exploit the vulnerabilities of Square and Multiply algorithm. Upon detection of SCA, we proposed random permutation function for cache mapping to hide the pattern of cache replacement policy. Additionally, we take the concept of hypothesis testing, deterministic formalism, and information theory to validate our approach.