Pseudorandom signals are periodic signals with many of the properties of random signals. They are deterministic, however, and so they can be used for repeated experiments. Their uses and their autocorrelation properties are discussed and then their power spectral density is considered. It is not flat. For a system with a large amount of noise, frequency response of the system can become lost in noise unless it is possible to average large numbers of blocks of data. This has resulted in an increasing interest in the design of signals with prescribed power spectral densities, in particular, in signals with a larger amount of power at a comparatively small number of frequencies. Two approaches to the design of such multifrequency signals are presented
Published in:
Multifrequency Testing for System Identification, IEE Colloquium on
Date of Conference: 8 Jun 1990