Delay and random scattering estimation with a band-limited signal: unconditonal CRB and MLE
Published in IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2026
Time-delay and source power estimation are fundamental tasks in a plethora of applications, in particular for remote sensing systems where the goal is to characterize the reflecting surface. Standard processing considers a Gaussian conditional signal model for which all the unknown parameters are assumed to be deterministic, but it may be more informative to consider a Gaussian random surface scattering, leading to an unconditional signal model. In this contribution, compact closed-form unconditional Cramér-Rao bound (CRB) expressions for delay and Gaussian source variance estimation are provided, considering a generic band-limited signal, as well as the corresponding maximum likelihood estimators (MLEs). The CRBs are expressed in terms of the signal samples, making it especially easy to use whatever baseband signal is considered. The results are validated with two representative band-limited signals to support the discussion.
Recommended citation: Filippo Torrisi, Corentin Lubeigt, Jordi Vilà-Valls, Eric Chaumette, "Delay and random scattering estimation with a band-limited signal: unconditonal CRB and MLE," IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Barcelona, Spain, May 2026.