KRIPTOSISTEM KUNCI ASIMETRI DAN PEMBANGKITAN KUNCI PUBLIK MENGGUNAKAN JARINGAN SYARAF TIRUAN RADIAL BASIS FUNCTION; ASYMMETRY CRYPTOSYSTEM AND GENERATING PUBLIC KEY USING RADIAL BASIS FUNCTION NEURAL NETWORK

HAYATY, NURUL, N

2016 | Disertasi | FMIPA

Security is very important part and can not be separated in a computer system and networks. The computer system is currently growing through network connections that are also expanding to make individuals and organizations more easily communicate and get information. Meanwhile, opportunities for parties who are not authorized to obtain confidential information becomes higher. One of the security methods is using cryptography. Key distribution symmetry of insecurity through the channels of communication gave rise to asymmetry of cryptography in 1976. Computational Cryptography has some similarities with neural network computing that is equally has the link between data input and data output. Neural network of Radial Bassis Function (RBFNN) is unique algorithm because the algorithm is hybrid method using combination of supervised and unsupervised learning methods. Plaintext represents the input layers, hidden layer represents the public key, the private key represents the weighs between the hidden layer to the output layer, and decrypted plaintext results represent the output layer. Based on the results of the research are constructed using the approach of neural network of Radial Basis Functions with function Gaussian activation can be used to encode the message and return it to its original form.

Kata Kunci : Cryptography, asymmetry, neural network, Radial Basis Function, Gaussian function.