A computationally-efficient, CRLB-achieving range estimation algorithm

Range splitting algorithms are used in radar signal processing to accurately estimate the range of a detected target by “splitting” the range bin and estimating where within the range bin the target is located. This paper presents a computationally-efficient range estimation algorithm that is empirically proven to achieve the Cramer-Rao lower bound (CRLB), while incuring minimal computational complexity. The algorithm calibrates the response in the mainlobe of the matched filter to incremental delays and is referred to as the calibration range estimation (CRE) algorithm. CRE then accurately estimates the target´s range from the received signal by linear interpolation of the calibration curve. We used both a stand-alone and a high-fidelity simulation to compare the range estimation error against the theoretical CRLB result. Using the stand-alone simulation, we showed that the CRE range estimate achieves the CRLB using at least twice the Nyquist sampling rate, and that it is moderately worse than the CRLB with the high-fidelity model. We also provide a brief discussion of the high-fidelity simulation, which captures antenna design and radar signal processing taper losses that the stand-alone simulation does not model.