Cellular automata-based noise generator (Q5926826)

The aim of the paper is to present a new analog noise generator that is built on the basis of digital techniques. The analog noise is obtained by digital low-pass filtering of a pseudo-random binary sequence (PRBS) extracted from an exhaustive cellular automaton (CA). A hybrid CA using 90 and 150 rules can be exhaustive. Cellular automata are fast, easily to be implemented, and the PRBSs produced by them display very good statistical characteristics. A 32-bit exhaustive CA has been built, and digital low-pass filtering of the PRBS waveform resulted in analog Gaussian noise with the spectrum depending on the clock frequency of the PRBS. The maximum circuit frequency achieved using standard HCMOS components on a breadboard was 20 MHz. This allowed a maximum noise bandwidth from 0 to 1 MHz. The proposed approach showed that digitally generated noise provides several advantages such as known spectrum and amplitude with adjustable bandwidth. Experimental results using the proposed technique are proved to support the theoretical setting.