Waveform diversity and design for agile sensing and environment characterization

In the last decade, agile systems have surfaced with the ability to change their transit waveform at each time step to adapt to their environment. In this dissertation, agile sensing techniques are proposed that characterize time-varying environments and adapt waveforms to exploit diversity in shallow water communication, radar, and multiple-input, multiple-output (MIMO) radar applications.;A normal-mode model for shallow water dispersive environments is investigated. A blind time-frequency processing technique is proposed to separate the normal-mode components by first approximating their time-frequency structure and then designing warping-based separation filters. Two types of receivers are developed and a matched waveform is designed to exploit the inherent model diversity and improve underwater communications.;For underwater tracking, a dynamic frequency-agile sensing algorithm that uses sequential Bayesian techniques is proposed to minimize the target state predicted mean-squared estimation error. The parameters of multiple acoustic sensors are designed to optimally estimate the target's location and velocity. The estimation is performed using sequential quadratic programming due to the high dimensionality of the sensor parameter grid search.;The Cramer-Rao lower bound (CRLB) for jointly estimating target attributes of moving targets are derived using MIMO radar, both with colocated and widely-separated antennas. For both types of systems, the CRLB is shown to depend on the parameters of the transmitted waveform at high signal-to-noise ratios. Waveform design for MIMO radar systems is investigated for dynamic target tracking by using the parameterized CRLB to minimize the target state estimation error. For a system with colocated antennas, frequency diversity is used to improve radar detection and estimation under varying conditions. The proposed system is a radar array with frequency-division multiplexing that can incorporate beamforming to design the transmission beam pattern.;In waveform-agile sensing applications, the environment characteristics need to be known as they can greatly affect the choice of waveform and sensing performance. Algorithms for estimating the spreading function of linear time-varying systems are proposed. The transmitted waveform is first designed to match the environment. Then, compressive sensing is used, based on the sparsity of the system characteristics, in order to reduce the number of measurements needed for the estimation.