An efficient parallel algorithm for secured data communications using RSA public key cryptography method

Public-key infrastructure based cryptographic algorithms are usually considered as slower than their corresponding symmetric key based algorithms due to their root in modular arithmetic. In the RSA public-key security algorithm, the encryption and decryption is entirely based on modular exponentiation and modular reduction which are performed on very large integers, typically 1024 bits. Due to this reason the sequential implementation of RSA becomes compute-intensive and takes lot of time and energy to execute. Moreover, it is very difficult to perform intense modular computations on very large integers because of the limitation in size of basic data types available with GCC infrastructure. In this paper, we are looking into the possibility of improving the performance of proposed parallel RSA algorithm by using two different techniques simultaneously, first implementing modular calculations on larger integers using GMP library and second by parallelizing it using OpenMP on the GCC infrastructure. We have also analyzed the performance gained by comparing the sequential version with the parallel versions of RSA running on the GCC infrastructure.