Nondata-aided hypothesis testing of PAM signals with energy detection

In this letter, we consider the problem of nondata-aided (NDA) detection of multilevel pulse-amplitude-modulated (PAM) signals combined with energy detection. We show that it is possible to estimate proper thresholds for the symbol decision without using redundant signals. Furthermore, we present an analytical method to evaluate the effect of pattern noise on the error performance of the NDA scheme with respect to the number of modulation levels and the convergence rate. With relatively low numbers of degrees of freedom and slow fading, which are typical for short-range systems using multiband modulation, the performance of the NDA scheme turns out to be asymptotically comparable to that of the optimal threshold selection.