Increasing the noise immunity of complex signal processing systems is the main problem in various areas of signal processing. At the present time there are many books and periodical articles devoted to signal detection, but many important problems remain to be solved. New approaches to complex problems allow us not only to summarize investigations, but also to improve the quality of signal detection in noise. This book is devoted to fundamental problems in the generalized approach to signal processing in noise based on a seemingly abstract idea: the introduction of an additional noise source that does not carry any information about the signal in order to improve the qualitative performance of complex signal processing systems. Theoretical and experimental studies carried out by the author lead to the conclusion that the proposed generalized approach to signal processing in noise allows us to formulate a decision-making rule based on the determi nation of the jointly sufficient statistics of the mean and variance of the likelihood function (or functional). Classical and modern signal detection theories allow us to define only the sufficient statistic of the mean of the likelihood function (or functional). The presence of additional information about the statistical characteristics of the like lihood function (or functional) leads to better-quality signal detection in comparison with the optimal signal detection algorithms of classical and modern theories.