SECURITY ANALYSIS OF A CHAOTIC RANDOM NUMBER GENERATOR BASED ON A FRACTIONAL-ORDER EXPONENTIAL JERK SYSTEM

In this study, the cryptanalysis of a chaotic random number generator (RNG) based on a fractional-order exponential jerk system (FOEJS) is discussed. In order to explain the target RNG's security flaws, an attack RNG is used. It is shown that using linear continuous feedback in a drive-response arrangement, consistent syn-chronization between the attack and target systems can be achieved if the RNG design is known and a state variable of the fractional-order chaotic system is observable for a time interval. Once the target and attack RNGs have synced, the attack system gener-ates the same output bit stream as the target RNG. Numerical simulations confirm the effectiveness of the attack approach. As an example demonstration, a RNG based on a fractional-order exponential jerk system is targeted in this article. Despite this specific target, the cryptanalysis method described in this paper can be applied to any RNGs based on fractional-order or integer-order chaotic systems, also based on continuous -time or discrete-time chaotic systems. As a result, this study highlights the security flaws associated with RNGs based on fractional-order chaotic systems, as well as the error of treating deterministic fractional-order chaos as a single entropy source for a RNG.