Data-Based Estimation of the Cramer-Rao Lower Bound for Direction-of-Arrival Estimation
B. Friedlander · 2022
The Cramer-Rao Bound (CRB) is a useful tool for studying many engineering problems involving parameter estimation. In order to make use of the CRB it needs to be calculated for the specific problem under consideration. This is typically done by deriving an explicit set of formulas which in many cases are quite complex. Alternatively, it is possible to estimate the CRB directly from data - either measured or computer generated. This paper describes a procedure for estimating the CRB in the Gaussian case and discusses its advantages and disadvantages. We note that the data-based CRB can be calculated even when an explicit mathematical model for the data is unavailable, or when there is significant mismatch between the assumed mathematical model and the real data.