Design of SMT Execution Port Side Channel Attack Detection Method Based on Ballet Algorithm

SCA (Side channel attack) often brings disastrous consequences to users or systems, so it is necessary to increase the side channel protection mechanism in hardware and software to improve the robustness of cryptographic algorithms. This paper mainly carries out related research work on SMT (Simultaneous multi-threading) port-side channel attack detection, focusing on the method of SMT port-side channel attack detection based on Ballet algorithm. The conversion activity of the application is inferred through the change of shared memory, and the activity of the foreground application is inferred according to the side channel resources exposed during the running of the application, such as memory occupation, CPU time consumption and network usage data. Memory occupation has the least influence on inferring foreground application activity, with an average of 1.58%. The detection scheme designed in this paper fully considers the characteristics of SMT execution port environment, and adopts a penetrating scanning method based on physical platform security configuration in the information collection stage to collect necessary event log information and security policy information, which reduces the workload of building a shared resource matrix.