International Journal of Engineering and Technology

Memetic algorithms have demonstrated their effectiveness to provide satisfactory solutions to combinatorial optimization problems reputedly hard.
In this paper, we present a new encryption system whose internal structure is essentially based on memetic algorithms. Our approach is based on the hybridization between an evolutionary algorithm based on solutions population and a local search adapted to the problem. The two methods are complementary, because the evolutionary algorithm can well sweep the search space, while the local search allows much more research in these areas to find the best solutions.
First, we will bring back the problem of encryption to a combinatorial optimization problem as in the Symmetric Encryption Evolutionary SEC. Then we will encode this problem in a specific way to bring us back to scheduling problems. Finally, after building the lists containing the different positions of the characters of the plaintext, we apply the memetic process on the order of these lists for maximum disorder. The performance criteria considered are the execution time and the convergence of the system. To validate the results found, we conducted a comparison to those found by the evolutionary algorithm.