The application of random number generator for interval modeling in multiple authentication process of online computer based training / Fakhrul Hazman Yusoff, Shaifizat Mansor and Syarifah Adilah Mohamed Yusoff

Online Computer-based Training (CBT) demand a method to ensure that the user maintain its present in-front of the computer throughout the session. One of the solutions is to force the application to invoke authentication process at regular interval. The challenge however is to ensure that the intervals are random and cannot be predicted by the user. This research compared three random number generators (RNG) that can be used to generate the interval for the authentication. The compared RNGs are Built-in VB, MRG32ka and Mersenne Twister. The comparison is to determine the suitability of the selected RNG for producing good authentication process for online CBT. The research also investigates the impact of using different seed source namely built-in clock and values derived from a Smart Card. The investigation is to determine whether usage of one of the seed source can enhance the quality of the RNG in term of better distribution and produced number sequence. The result of this research showed that although Mersenne Twister claimed that it can produce the longest sequence among the three RNGs, MRG32ka and Built-n VB RNG turn out to be a better RNG to be implemented for multiple authentications for online CBT. Meanwhile usage of different seed source did not contribute much to the quality of the RNG.