A Chaotic Non-Linear Pseudo Random Number Generator (CNPRNG) Design for High-Security Applications

This paper presents a novel Chaotic Non-linear Pseudorandom Number Generator (CNPRNG) design, integrating a linear feedback shift register (LFSR) with a modified non-linear chaotic Cellular Automaton (CA) using Rule 30 to enhance randomness quality and thus security for lightweight devices. The proposed hybrid architecture addresses the inherent predictability issues associated with traditional LFSR-based generators by introducing complexity through non-linear dynamics. Synthesized on a Spartan 7 FPGA, the design demonstrates superior performance in terms of resource efficiency, consuming only 42 LUTs and 64 registers while operating at 100 MHz and maintaining low power consumption at 127 mW. Rigorous testing through NIST SP800-22, and AIS-31 test suites confirms its high entropy and robust randomness, qualifying it for high-security applications. The CNPRNG's effectiveness is further affirmed by achieving minimal bias and autocorrelation characteristics among the generated bitstream. Finally, the proposed method was synthesized for three technology nodes in the Synopsis Design Compiler (SDC) to verify the scalability and resource utilization.