Title :
Estimating the number of signals and signal resolution
Author_Institution :
Dept. of Math. & Stat., Louisiana Tech. Univ., Ruston, LA
fDate :
8/1/1998 12:00:00 AM
Abstract :
The nested likelihood ratios are shown to be asymptotically independent and sufficient for estimating the number of signals and to implement resolutions of superimposed signals into individual signals. As a result, estimating the number of signals can be reduced to a sequence of independent and simple tests. The procedure suggests an engineering sense that estimating the number of signals is based on complete resolutions of multiple signals. Simulations for estimating the number of overlapping echoes are given for illustration
Keywords :
maximum likelihood estimation; sequences; signal reconstruction; signal resolution; D-sequence; multiple signals; nested likelihood ratios; overlapping echoes; signal numbers estimation; signal resolution; superimposed signals; Gaussian noise; Mathematics; Maximum likelihood estimation; Signal processing; Signal processing algorithms; Signal resolution; Signal to noise ratio; Statistical distributions; Statistics; Testing;
Journal_Title :
Signal Processing, IEEE Transactions on
