Modernizing Cryptography: High-Throughput Reseeding-Mixing PRNG in the Design and Implementation of a Cutting-Edge Algorithm

Abstract

The Pseudorandom Number Generator (PRNG) plays a vital role in number sequence generation and is highly regarded for its swift production of pseudorandom numbers. Within cryptography, the Reseeding-Mixing PRNG (RM-PRNG) algorithm is extensively employed as a key generation mechanism, ensuring the secure transmission of large datasets without vulnerability to interruptions or hacking endeavors. This thesis introduces an innovative reseeding mixing method designed to extend the system period length and improve the statistical properties of the PRNG. The reseeding technique eliminates brief periods within the digitized logistic map, while the mixing method prolongs the system period length to an impressive 2253 through "XOring" with a DX generator. The resulting RM-PRNG attains a remarkable throughput rate of 6.4 Gb/s, surpassing other nonlinear PRNGs. The output sequence of the RM-PRNG serves as a key for encryption and decryption modules. The complete design is meticulously detailed in Verilog Hardware Description Language (HDL) and thoroughly verified for functionality using Xilinx ISE 9.1i software.