Analysis of Key Generataion Using Neural Networks in Mobile Adhoc Networks
MANETs are self-conﬁguring systems of versatile nodes not including the nearness of static foundation. MANETs need more security due to its characteristics and for securing data cryptographic technique are required. These techniques require the keys for securing data as private and public keys and this key management is a typical sensitive issue. Neural network based techniques provide the untraditional methods for it. In the present paper, we have presented the simulation of keys and performed different experiments with a forty-bit key set by taking different number of neurons in the hidden layers of neural network. A back propagation learning algorithm which is based on feed forward neural network pattern is designed. This works has been implemented on different sizes of hidden layers, and results obtained have been presented. The results have helped us to conclude that the convergence rate is high in back propagation algorithm. The three layered network architecture is taken, larger layered architecture may lead to more time and space complexity.
Keywords- MANETs, Cryptography, Key Management, Neural Networks, and ANN Model.