Signal Detection for Stochastic Processes with Stationary Independent Symmetric Increments.

摘要

For processes with stationary independent symmetric increments, several one-sample and 2-sample detection problems are treated. The statistical tools are based on minimal sufficient statistics and maximal statistical noise and employ rank-sum, sign, signed-rank and Kolmogorov-Smirnov statistics. Relevant point and interval estimation techniques are presented as are numerical examples illustrating the techniques. (Author)