For the full lecture visit https://www.drnirregev.com/lecture-peak-location-estimation

Lecture: Peak Interpolation in Radar and Lidar (Snippet): Quadratic interpolation scheme for post-FFT / post matched-filter peak location estimation is heavily used in the industry. My advice: DON'T use it! As we will learn in this lecture, the quadratic interpolator/estimator (orange curve) has a bi-modal error distribution and very large RMSE compared to the Cramer-Rao lower bound (CRLB, blue curve).

In contrast, other estimators provide far better accuracy across the range of possible peak locations. While the CRLB is a theoretical limit, many practical estimators exist that approach its performance as we will see in the lecture.

When selecting a peak location estimator, RMSE is a key metric to consider, but one should also evaluate computational complexity, bias, and consistency. The right choice depends on the specific system requirements and constraints.

Join my paid lecture on how to efficiently estimate the peak location post-FFT. In this lecture, we will dive into the mathematical details of various estimators, including the quadratic, MATLAB code will be provided to test out each approach. You will leave this lecture with a deep understanding of this problem and with an arsenal of tools to take and implement into your DSP pipeline.

To get the most out of this lecture, attendees should be comfortable with radar/DSP fundamentals, basic estimation theory, and MATLAB/Python programming. Don't miss this opportunity to master a critical skill for any radar signal processing engineer or researcher!