System Security Employing Odor Identification with Neural Networks

Security technique that make use of odors as passwords is presented in this work. The implemented system employs odor keys to allow access to secure systems. The used key combinations are selected randomly before each time a system needs to be accessed, thus, there is no way of being emulated beforehand. The work proved that odor keys can be used to access secure systems using artificial neural networks that have the ability to learn and discriminate between odors using odor arrays. The presented work also introduces the concept of data packetizing with the duration of response of the used electronic nose (E-Nose) as part of the header, with the number of counted pulses representing the E-Nose response as the frame data. The header is used as part of the control information that checks the correctness of the contents of the represented data frame.