| Cloud computing is a paradigm that brings about profitability and conve-nience to name few,by utilizing resources fully.Scenarios where two or more users a located on the same CPU but on different cores in different virtual machines are common.Although there are software-based security and privacy measures to protect users in the cloud,unfortunately these measures are not perfect so it is possible for malicious users to stealthily steal private information from other users.Detection methods have been presented that stop cache side-channel at-tacks which cause significantly more cache-hits and cache-misses.Using Hardware Performance Counters(HPCs)we observe that our attack is stealthy compared to a benign application the Firefox Web Browser in terms of cache activity.With our attack,the attacker can stealthily steal private information about the last visited websites of the victim,thereby obtaining information that can be used to profile the victim without any authorization or knowledge of the victim.Our attack works in a scenario where both the attacker and victim are located on the same CPU but on different cores,in different virtual machines and have the same application installed.Our results suggest that the shared memory capability may be vulnerable to cache side-channel attacks and that there is need to divert some attention to mitigating cache side-channel attacks in the cloud. |
