A new class of discrete modular memristors and application in chaotic systems

With the development of memristor, memristive chaotic systems have attracted attention, but many of them have divergence problems. To solve this problem, in this paper, a universal discrete modular memristor (DMM) model is proposed to overcome the problem of system divergence. Based on this model, a class of novel memristors are designed, and it is proved that they satisfy three fingerprint characteristics by theoretical verification and experimental simulation. Meanwhile, taking a one-dimensional (1D) chaotic map and a two-dimensional (2D) chaotic map as seed chaotic maps, two memristive chaotic maps are generated, and their dynamics are analyzed by phase diagram, Lyapunov exponent and complexity. Experimental results show that these memristive maps have good ergodicity, large chaotic range, hyperchaotic behavior and high complexity. In addition, these memristive maps are implemented on DSP platform, and applied in pseudorandom number generator (PRNG), which further validates their application potential.