| Summary: | 碩士 === 國立臺灣海洋大學 === 電機工程學系 === 96 === The thesis focuses on enhancing security during the data transmission involed with Radio Frequency Identification (RFID). A novel mechanism is presented to deal with transmission of product daya over wireless channel. In this mechanism, product data (i.e. sensor readings, invoice, and production traceability) is acquired through RFID incorporated with a wireless sensory network (WSN), an encryption system is then employed to protect the product data.
In the thesis, two separate (but with the same architecture) neural networks are disposed at both receiver and transmitter ends. At the stage of producing key, the receiver and transmitter will exchange their network outputs, and the two outputs are compared to determine if further training is needed. The resulting network connection weights are used as the key string. The key and RFID tag data are used to find in the back-end database the confidential data and send it to an encryption mechanism for completing cryptography, and finally the correctness is verified by an authentication process. Because a random number is used to update the tag data every time after a data transmission is complteted, data integrity and confidentiality are ensured while replay attack is prevented. In order to verify its effectiveness in integrity, confidentiality, and non-repudiation, our mechanism is compared to others by applying security analysis [1-3] to test the six well-known malicious attacking modes.
|
|---|
