Information Security for Products Record in Wireless Sensory System

Abstract

This paper utilizes neural network’s capability in learning non-linear input/output function mapping to achieve irreversibility and unpredictability requirements in information security. The work focuses on (secret) key and key exchange security system in coordination with a random scheme to produce symmetric keys. An information security system is constructed by using the key exchange security mechanism, it applicability is illustrated with a products trace backing system. In formation transmission security is accomplished by incorporating key generated by neural network, SHA-256, AES and XOR. The beauty of this approach is threefold: (1)the key is produced faster than [18], and needs not to be transmitted, thus greatly enhancing security (2) it can deal with short (e.g. RFID) or lengthy data strings (e.g. WiMax), as fast data encryption and identification are ensured (3) capable of solving the collision problem in ordinary hashing function. Another contribution of this work includes the improvement of the method [18] of using neural network in acquiring public key.