Abstract: In digital communication systems, the signal-to-noise ratio is an important input parameter for a variety of equalization and decoding algorithms. In typical applications such as transmission over fading channels both, current signal and current noise power are unknown. An algorithm is presented which allows estimation of signal and noise. The entire a-priori information necessary are fundamental properties of the probability density functions of signal and noise. The dependence of the quality of the estimation on the ensemble size used and on the actual signal-to-noise ratio is discussed. Simulation results for continuous as well as for blockwise transmission show that a small error of the estimate is achieved for rather small ensemble sizes.
Key words: Higher order statistic, digital communication system, signal to noise ratio, probability density function, fading channels, trellis coded modulation, PSK.