Adaptive waveform design for SAR in a crowded spectrum
This thesis concerns the development of an adaptive waveform design scheme for synthetic aperture radar (SAR) to support its operation in the increasingly crowded radio frequency (RF) spectrum, focusing on mitigating the effects of external RF interference. The RF spectrum is a finite resource and the rapid expansion of the telecommunications industry has seen radar users face a significant restriction in the range of available operational frequencies. This crowded spectrum scenario leads to increased likelihood of RF interference either due to energy leakage from neighbouring spectral users or from unlicensed transmitters. SAR is a wide bandwidth radar imaging mode which exploits the motion of the radar platform to form an image using multiple one dimensional profiles of the scene of interest known as the range profile. Due to its wideband nature, SAR is particularly vulnerable to RF interference which causes image impairments and overall reduction in quality. Altering the approach for radar energy transmission across the RF spectrum is now imperative to continue effective operation. Adaptive waveforms have recently become feasible for implementation and offer the much needed flexibility in the choice and control over radar transmission. However, there is a critically small processing time frame between waveform reception and transmission, which necessitates the use of computationally efficient processing algorithms to use adaptivity effectively. This simulation-based study provides a first look at adaptive waveform design for SAR to mitigate the detrimental effects of RF interference on a pulse-to-pulse basis. Standard SAR systems rely on a fixed waveform processing format on reception which restricts its potential to reap the benefits of adaptive waveform design. Firstly, to support waveform design for SAR, system identification techniques are applied to construct an alternative receive processing method which allows flexibility in waveform type. This leads to the main contribution of the thesis which is the formation of an adaptive spectral waveform design scheme. A computationally efficient closed-form expression for the waveform spectrum that minimizes the error in the estimate of the SAR range profile on a pulse to pulse basis is derived. The range profile and the spectrum of the interference are estimated at each pulse. The interference estimate is then used to redesign the proceeding waveform for estimation of the range profile at the next radar platform position. The solution necessitates that the energy is spread across the spectrum such that it competes with the interferer. The scenario where the waveform admits gaps in the spectrum in order to mitigate the effects of the interference is also detailed and is the secondary major thesis contribution. A series of test SAR images demonstrate the efficacy of these techniques and yield reduced interference effects compared to the standard SAR waveform.