Efficient Implementation of the Classic McEliece on ARMv8 Processors

Classic McEliece is a Code-based Key Encapsulation Mechanisms (KEM) and one of the candidate algorithms in the NIST PQC competition. Based on the McEliece cryptosystem developed in 1978, this system relies on the Niederreiter variant of McEliece. It consists of three phases: Key Generation, Encapsulation, and Decapsulation. In this paper, we propose an optimized implementation of the internal multiplication operations of Classic McEliece on the ARMv8 processor. We utilize parallel computing techniques using vector registers and vector instructions of the ARMv8 processor. We specifically focus on optimizing the multiplication operation, which is a major contributor to the overall execution time of the Classic McEliece algorithm, by leveraging the commutative property and implementing an parallelization technique. As a result, our approach achieves a maximum performance improvement of \(2.82\times \) compared to the reference implementation in the multiplication operation.