International Journal on Advanced Science, Engineering and Information Technology, Vol. 8 (2018) No. 4, pages: 1327-1332, DOI:10.18517/ijaseit.8.4.5738

A Study on Machine Learning Based Light Weight Authentication Vector

Do-Hyeon Choi, Jung-Oh Park


Artificial Intelligence area has been rapidly advanced around the global companies such as Google, Amazon, IBM and so on. In addition, it is anticipated to facilitate the innovation in a variety of industries in the future. AI provides us with convenience in our lives, on the other hand, the valuable information on the subjects that utilize this has the potential to be exposed at anytime and anywhere. In the next advancement of AI area, the technical developments of the new security are required other than the existing methods. Generation and validation methods of light-weight authentication vector are suggested in this study to be used in many areas as an expanded security function. Upon the results of the capacity analysis, it was verified that efficient and safe security function could be performed using the existing machine learning algorithm. Authentication vector is designed to insert the encrypted data as variable according to the change of time. The security function was performed by comparing coordinate distance values within the authentication vector, and the internal structure was verified to optimize the performance cost required for data reverse search.


Artificial Intelligence; Machine Learning; Authentication Vector; Virtualization; Next Generation Security

Viewed: 543 times (since abstract online)

cite this paper     download