A First Look at Combining Pulse Compression and Range Oversampling

Weather radars based on low-power solid-state transmitters typically use pulse compression as a means to achieve the required sensitivity to detect weaker returns. Pulse compression works by lengthening the transmitted pulse and modulating the transmitted waveform. A longer pulse leads to increased sensitivity due to the higher average power being transmitted, and frequency or phase modulation can be used to recover the original range resolution after processing. As such, pulse compression waveforms require a larger transmission bandwidth and result in an extended blind range close to the radar, which is typically mitigated by the addition of a fill-in pulse. Although the resulting range weighting function has a main lobe with the desired width, it contains sidelobes that extend a few kilometers in range. Range sidelobes can impact the range resolution of the radar and are typically reduced by employing non-linear frequency-modulation transmission schemes at the price of an increased transmission bandwidth compared to conventional linear-frequency-modulation waveforms. Range oversampling processing has been used on radars with conventional high-power transmitters to reduce observation times without increasing the variance of estimates, but range oversampling and pulse compression have yet to be combined. Range oversampling processing consists of sampling the received signals at a rate faster than the inverse of the transmitted pulse width thus producing complex voltages with a range correlation that depends on the modified pulse shape (the modified pulse is the convolution of the transmitted pulse envelope and the receiver filter impulse response, which would include the effects of pulse compression). This a-priori information about the range correlation is used to devise transformations to decorrelate sets of range-oversampled signals from which auto- and cross-covariances are estimated. These are averaged to match conventional range sampling and are ultimately used to obtain more precise estimates of all radar variables. In this work, we take a first look at combining pulse compression and range oversampling, which is applicable to the design of any weather radar system that uses pulse compression for enhanced sensitivity and needs range-oversampling processing for faster updates such as the upcoming Multifunction Phased Array Radar (MPAR) Advanced Technology Demonstrator (ATD). Because both techniques involve processing in range time, it is important to understand how one affects the other. Since pulse compression determines the modified pulse, the performance of range oversampling is intimately tied to it. Our study focuses on providing a framework to quantify the performance of the combined techniques and on exploring possible complications for a practical implementation on the MPAR ATD.