Maximum A Posteriori Likelihood Detection Using Lempel-Ziv Algorithm

We present an efficient, semi-optimum data-processing scheme for detecting a deterministic signal in white Gaussian noise and a random transient disturbance which occurs unpredictably and infrequently. Data are processed in two stages. The first consists of binary quantization of the matched-filter output and parsing. The latter is a simpler version of the Lampel-Ziv algorithm. We use it to determine whether or not the disturbance has occurred during the observation interval. If it is decided that the disturbance has not occurred, then the second stage is simple to threshold the matched-filter output as in the classical detection problem.