Spectrally constrained active sensing: Waveform and receiver filter design

Traditional active sensing waveform and receiver filter design has always assumed that a continuous block of spectrum will be allocated for use. Due to growing demand for bandwidth among many different fields large continuous blocks of spectrum are not available. However, large bandwidths are required to meet the system requirements for high performance active sensing systems. A system that has to operate under such a constraint would be a spectrally constrained active sensing system. Traditional waveform design and receiver filters are not optimal in this scenario. In this work, we present a waveform design algorithm called SHAPE for the spectrally constrained problem. We also present a receiver filter for processing spectrally constrained linear frequency modulated waveforms in a synthetic aperture radar (SAR) application. Finally, we present a novel method of designing spectrally constrained waveform sets for application in multiple-input multiple-output SAR.;This work begins by examining the fundamentals of active sensing in the introduction. In active sensing the common goals are to estimate the range and velocity of objects in a sensor's field of view. This is accomplished by measuring the echoed returns of a known transmitted signal and any frequency shifts applied to that signal by an object's velocity (which is caused by the Doppler effect). The choice of the transmitted signal is crucial because it determines our ability to resolve multiple objects, any interference due to neighbouring objects, and how the signal reacts to Doppler shifts at the receiver.;A very good design choice is to use a signal that has a very large bandwidth and has energy uniformly distributed across the entire bandwidth. However, due to the limited nature of the frequency spectrum and the large bandwidth requirements we may not be able to meet all design requirements because other users will be occupying specific bands. If we radiate energy without considering the interference we will cause, then it is an illegal activity, and it could have catastrophic effects such as disrupting crucial aviation navigation equipment. To overcome this we consider spectrally constrained active sensing. The first problem we address in spectrally constrained active sensing is how to design a probing waveform that does not cause interference in some given bands. A design algorithm called SHAPE is presented for developing such a probing signal.;The effects of a spectrally constrained probing signal may have unwanted consequences in active sensing. In certain applications such as synthetic aperture radar, it will result in a missing data problem. The missing data will degrade the SAR's output often to an unusable level. To overcome this we propose to use the missing data iterative adaptive approach (MIAA) to estimate the missing data. However, SAR data is often very large and the traditional MIAA algorithm is not optimized for computational efficiency. A fast algorithm was proposed, but it is only efficient when over 50% of the data is missing. For this reason, we developed a new fast MIAA algorithm that is computationally efficient when the missing data is less than 50%. We demonstrate the effectiveness of the algorithm on some simulated missing data SAR examples.;We also consider the case of spectrally constrained waveform design for SAR and multiple-input, multiple-output (MIMO) SAR ground moving target indicator (GMTI). Here we show that by using Kronecker waveforms (one waveform embedded in another), it is possible to generate spectrally constrained waveforms that have low correlation zones. Furthermore, by using these Kronecker waveforms we are given control of the cross-correlation between the waveforms in the set. Using this approach we present a novel method for generating spectrally constrained waveform sets, which has not been explored in the recent literature. We demonstrate the efficacy of these waveforms using SAR and MIMO SAR GMTI simulations.;This work is concluded by examining the possibilities and future directions of the spectrally constrained active sensing approaches described within. A major milestone is to perform tests using updated software defined radars. By doing this we can better understand the limitations of the designed waveforms and what parameters are more important than others. Here we could also understand how the non-linear effects of the transmitter and receiver front-end distort our signal and how to account for that in the design process. Finally, we would like to study improved methods of designing the receiver to improve detection and performance of these signals in extremely cluttered environments. All of these are critical to developing active sensing systems that can meet the upcoming stringent spectral requirements of the future.