Canonical non-Gaussian noise models: Their implications for measurement and for prediction of receiver performance

The paper provides a summary of principal noise model results and considers their major implications for measurement and performance prediction and evaluation. The discussion is illustrated by a concise review of model statistics and methods of estimating the model parameters, including the effects on accuracy of finite data samples. A variety of signal detection problems are used to illustrate the potential improvement possible when optimum algorithms are employed, vis-a-vis most of the currently used systems, which are conventionally optimized for Gaussian interference. Conditions for realizing such gains are also given, and the principal key properties of these new, general models are discussed.