Adaptive Noise Estimation for Dual-Polarized Phased Array Weather Radars

Noise floor estimation is an important aspect of weather signal processing. An estimate of the radar noise floor is used in the computation of calibrated reflectivity and in data thresholding algorithms. Due to changing environmental conditions and external interference, however, the noise floor presented to the radar can vary with time and pointing angle. If the noise estimates cannot respond to changes in the radar noise floor, this can introduce a bias in weather product estimates. To reduce this bias, an accurate and up-to-date estimation of the radar noise floor is required. Traditional methods of noise estimation use dedicated noise dwells inserted into the radar timeline at periodic intervals. Due to the advent of affordable phased array technology, more flexible and adaptive noise estimation techniques can be used that take advantage of the electronic scanning capabilities. This paper presents a new method for estimating the radar noise floor accurately in phased array radars, and compares the method to traditional techniques.

Phased array radars offer increased flexibility in adaptive waveform selection and receive beam shaping, but these capabilities can affect the radar noise floor of the system. This new noise estimation technique efficiently estimates the noise floor in real-time by injecting short bursts of transmit-disabled pulses into the radar timeline at each radial. The algorithm dynamically adjusts the number of noise pulses per estimate, based on the revisit rate at the pointing angle, the waveform being transmitted, and the receive array taper used. In addition, a history of noise estimates at each angle is maintained to track changes in the noise floor and detect the presence of external RF interference.

This paper presents simulation and live data results that show the improvement of the new method over traditional techniques. Data will be presented that compares the accuracy and radar timeline usage of the new method to traditional methods used on operational systems. Results will demonstrate the ability of the algorithm to correctly detect the presence of RF interference and estimate the noise floor for different waveforms and receive tapers. This noise estimation technique allows for flexible radar waveform design and receive beam shaping in phased array weather radars.