Side Channel Analysis Library
The Side Channel Analysis Tool is an open source tool written in C++ that allows the user to recover encryption keys from leakage data obtained through some form of Side Channel Analysis. Currently only Correlation Power Analysis (CPA) for the AES-128 encryption algorithm is supported but we hope to continue supporting and expanding the range of analysis and encryption algorithms for the future.
The Side Channel Analysis Tool has several dependencies that must be installed before use:
- CMake - CMake is a family of tools used to simplify the compilation process. The minimum required version of CMake for the SCA Tool is 2.8
- OpenMP - OpenMP is an API that allows for multithreading and is used througout the tool to speed certain processes.
After both dependencies are installed the tool can be built. In the top level directory there is an empty directory labelled 'build'. This is the recommended location to launch the CMake build process to avoid cluttering the source files. Note that building in a different location will cause problems for the GPU algorithm since the executable needs to know the location of the OpenCL kernel at runtime. The following commands will build the tool assuming the user is in the top-level directory of the tool:
After step 2, cmake will configure the makefiles required for building the project. At this point, cmake will search for the OpenCL shared library as well as the necessary OpenCL headers required to compile the GPU-based algorithms. If these are not found, the tool will only build the CPU-based algorithms. After step 3, the user should have an executable file labelled 'sca' which will be used to invoke the analysis algorithms.
Once built, the Side Channel Anlysis tool has three different command line flags available:
-d- The argument following this flag specifies the data file with the power leakage traces. This argument is mandatory.
-t- The argument following this flag specifies the ciphertext file associated with the provided data file. This argument is mandatory.
-p- When this flag is specified the GPU-based CPA algorithm will be invoked instead of the CPU-based algorithm. This can provide a performance increase for large sets of data.
Usage of the CPU-based CPA algorithm:
./sca -d <path/to/power/traces> -t <path/to/ciphertext>
Usage of the GPU-based CPA algorithm:
./sca -d <path/to/power/traces> -t <path/to/ciphertext> -p
In the directory labeled 'data', 3000 power traces collected from a SASEBO-GII FPGA are provided to test the Side Channel Analysis Tool.
If you have any question, please contact
The software is available to the public for free. If you use them in your work, we request you to kindly acknowledge us by citing our website with the title of “Northeastern University TeSCASE dataset” and the URL of “https://chest.coe.neu.edu/.”