Hyperjerk multiscroll oscillators with megastability: Analysis, FPGA implementation and a novel ANN-ring-based True Random Number Generator
Özet
Metastable oscillators are a special case of chaotic attractors with infinite coexisting attractors. In this paper we propose a class of hyperjerk chaotic oscillators which exhibits megastability with and without forcing term. As the systems have infinite equilibrium points, all these coexisting attractors falls in the special category of hidden attractors. We have considered the unforced hyperjerk system to show the multiscrolls and also to investigate the system bifurcations. Then, the proposed hyperjerk multiscroll chaotic system has been designed as ANN-based on FPGA. TanSig activation function, used in the design of ANN-based chaotic oscillator, has been modeled using CORDIC-LUT approximation. To improve the chaos-based engineering applications, this paper presents a new dual entropy core hybrid TRNG having high speed by using the structures of ANN-based chaotic system and ring oscillator. The designs have been coded in VHDL with respect to IEEE-754-1985 number standard. The implemented new proposed TRNG unit has been synthesized for Virtex-6 FPGA chip. After the Place-Route process, FPGA chip statistics and maximum operating frequency have been given as 167.4 MHz. 1 million bit stream generated by the proposed TRNG has been subjected to NIST-800-22 randomness tests and it has successfully passed all of the randomness tests. (C) 2019 Elsevier GmbH. All rights reserved.
