Variability analysis of tent map-based chaotic-map truly random number generators

Although discrete-time chaotic-map TRNGs, in principle, are capable of generating unbiased and uncorrelated random bits, implementation variations deteriorate the quality of the output sequence as a result of variation of the chaotic map parameters (which result in a non-ideal map). In this paper, we introduce an analytical technique to investigate the impact of process variations on the randomness of the output sequence generated using tent map-based discrete-time chaotic map truly random number generators (TRNGs). We validate our results against the probability density function (pdf) estimated using extensive Monte-Carlo simulations applied to the non-ideal map characteristics. The simulation results suggest that the estimation of pdf using our simplified theoretical method is in good agreement with that of the Monte-Carlo simulations. Finally, the quality of the random bits have also been evaluated using the entropy measure.