MSE lower bound for compressive parameter estimation: From Cramér-Rao to Ziv-Zakai

Target parameter estimation is one of the fundamental functions of radar/sensing systems. The accuracy limit of parameter estimation revealed by the mean square error lower bound also received extensive attention and application from researchers. Therein, the Cramer-Rao bound (CRB) is well-accepted due to the easy evaluation, where, however, the locality makes it only tight in high SNR situation. With the development of the signal processing, target parameter estimation in moderate even low SNR also becomes an important topic. However, it is difficult for the widely used CRB to tightly evaluate the accuracy of the estimator, which, leads to an urgent of introducing a tighter MSE lower bound over a wide range of SNR. In this talk, I will briefly review the existing MSE lower bounds, and introduce a global tight Ziv-Zakai bound. Then, I will introduce our recent work on Ziv-Zakai bound for time delay estimation and DOA estimation.