This paper presents a new approach based on artificial neural networks (ANNs) for data security in electronic communication. Two ANN modules were used for encryption and decryption. ANN modules were trained with Levenberg-Marquardt learning algorithm. Character based learning and test processes were followed. A program developed in Delphi achieves the processes automatically. The results have shown that the cryptosystem based on ANNs is very successful and also provides more security with the help of adding an extra module. It is thought that the study presented in this work will enable new directions to the applications because of real-time application of ANNs.

This paper presents a new approach based on artificial neural networks (ANNs) for data security in electronic communication. Two ANN modules were used for encryption and decryption. ANN modules were trained with LevenbergMarquardt learning algorithm. Character based learning and test processes were followed. A program developed in Delphi achieves the processes automatically. The results have shown that the cryptosystem based on ANNs is very successful and also provides more security with the help of adding an extra module. It is thought that the study presented in this work will enable new directions to the applications because of real-time application of ANNs