Techniques for homodyne dechirp-on-receive linearly frequency modulated radar

Abstract

This thesis presents work done to extend and improve the operation of homodyne dechirp-on-receive linearly frequency modulated radars.

First, an investigation of the effect of common phase errors on the point response function of the radar is described. The dependence on the window function of the degradation due to phase errors is investigated, and a simple, precise, and general approach for calculating the degraded Point Spread Function (PSF) is described and demonstrated. This method is shown to be particularly useful when investigating the effect of chirp nonlinearity on the PSF.

Next, a method for focussing range profiles that are degraded by chirp nonlinearity is described. This method is based on two established methods, the Phase Gradient Algorithm (PGA) and a time-domain re-sampling technique. The technique is entirely hardware independent, allowing any homodyne dechirp-on-receive linearly frequency modulated radar to be focussed. Where suitable archive signal data exists, focussed imagery can even be produced from radars that no longer exist. The complete algorithm and details of the implementation are described, and the technique is demonstrated on three representative radar cases: extreme chirp nonlinearity, typical chirp nonlinearity, and a retrospective case. In all of the cases, it was shown that the PSF was dramatically improved.

A technique based on down conversion by aliasing for reducing the required sampling rate is described, and a simple technique for calculating suitable sampling rates is presented. This method is demonstrated for a typical application in which sampling rate reduction might be required, namely Moving Target Indication (MTI). The MTI application is described and quantified, including a simple technique for choosing suitable radar operation parameters. The MTI technique with subsampling was demonstrated in software simulations and in a simple radar experiment.

A Synthetic Aperture Radar (SAR) test bench for researching component performance and scatterer properties in the context of SAR was developed. An appropriate image formation processing algorithm was found and modified to better suit the task of a short data collection baseline and drifting centre frequencies, both of which are present in the test bench situation. Software was written to collect data, to control the hardware, and to process the signals into SAR images. A data simulator was written to test the image formation algorithm implementation; it also served as a useful tool for investigating the effect of signal errors on the quality of the resultant SAR imagery. A suitable oscillator was chosen for the task, based on phase noise and centre frequency stability considerations, both of which are quantified and discussed. Preliminary SAR imagery was produced, indicating that the system operates correctly and in agreement with comparable systems.