A broad-spectrum pseudorandom Gaussian noise generator

A new technique is described for filtering a pseudorandom binary sequence to produce pseudorandom noise that has a near-Gaussian amplitude probability density function. Its principal advantages are that the half-power bandwidth of the noise is approximately eight times that obtained with existing techniques, and the desirable autocorrelation properties of the binary sequence are retained. A practical design for a noise generator using the new technique is described. The results of simulation are given to substantiate the advantages claimed.