Abstract:
Easily realizable robust algorithms are proposed which detect deterministic signals against the background of noise: they rely on robust criteria for the signals to approach the process. Extensions to the continuous and multioption case are described.