Extended HTLS methods for parameter estimation of multiple datasets modeled as sums of exponentials

摘要

In subspace-based parameter estimation of a single signal modeledas a sum of exponentially damped sinusoids, a variant of Kung's method,called HTLS in nuclear magnetic resonance spectroscopy, achieves bestresolution and accuracy. When more than one signal is to be quantified,the HTLS method can be extended in several ways. This paper presentsseveral extended HTLS methods quantifying multiple signals and comparesthese methods in different scenarios by means of a simulation study tofind out which and when is best suited