Speedup Higher-Order Masking of AES sing Normal Basis and SIMD

This paper introduces a new efficient technique to speed up the higher-order masking scheme of the advanced encryption standard (AES). The efficiency of the proposed technique is based on using normal basis in the square operation, using Karnaugh map to lower the required extensive calculations, and using single instruction multiple data (SIMD) technology to perform parallel processing at data level. A Linux machine with an Intel processor that supports SSSE3 instructions is used to evaluate this technique. The technique reduces the encryption cycles by 32% over the second order and 27.1% over the third order masking scheme reported in the literature. We achieved much more practical and attractive results than the reported in the literature.