ADAPTIVE WAVEFORM DESIGN FRAMEWORK FOR MIMO RADAR UNDER PRACTICAL CONSTRAINTS

The recent developments in radar technology - powerful signal processors, increased modulation bandwidth and access to higher carrier frequencies - offers enhanced flexi- bility in waveform design and receiver processing. This provides additional degrees of freedom in the signal design and processing, thereby offering additional avenues to im- plement interference mitigation. The radar environment is dynamic in general, with the inhomogeneous interference sources changing rapidly both in space and time. In this context, an adaptive waveform and adaptive receiver design for Multiple-Input-Multiple- Output (MIMO) radar system is a promising way forward towards dynamic interference mitigation. Even-though the technology offers flexibility, the need to commercialize radar elements imposes certain constraints on the platform to ensure commercial viability. In this context, the transmitted waveform has to satisfy practical design constraints imposed by the hardware including discrete phase modulation and limited number of processing chains. These coupled with the dynamic scenarios warrants a rapid signal adaptation with enhanced performance while satisfying the design constraints. Motivated by the aforementioned requirements, the thesis proposes a general framework for MIMO radar signal adaptation under practical design constraints. The transmit antennas are restricted to operate in a multiplex mode, where a fewer number of pro- cessing chains are multiplexed across an arbitrary number of transmit antennas. Each of these chains, also referred to as channels, have the capability to modulate the phase of a traditional radar pulse in discrete steps. Further, the modulation is assumed to be in the slow time domain (inter-pulse); such a phase modulation results in benign requirements on the platform. Furthermore, the antennas are assumed to be mounted uniformly in a way that the virtual MIMO paradigm for maximum angular resolution is satisfied. The slow time modulation naturally results in in an angle-Doppler coupling; this issue is addressed by phase center motion (PCM) techniques, where nonlinear and random PCM techniques for mitigating angle-Doppler coupling are proposed. While the PCM techniques provide orthogonal signals, a transmit beamforming approach is also consid- ered to exploit the salient features of MIMO and phased array radars. Towards this, an approach based on block circulant decomposition for the slow-time modulation is proposed to generate a particular beam shape while minimizing the cross-correlation between transmitted signals, such that the virtual MIMO paradigm is satisfied. The thesis formulates the radiation pattern design as a dictionary based convex optimization and proposes closed-form signal design solutions for particular configuration of channels, discrete phase stages and transmit antenna elements. The beampattern design is then elegantly combined with the PCM approach to reduce Doppler ambiguity while sup- pressing angle-Doppler coupling. The proposed waveform design methodology is shown to be amenable to fast adaptation. Further, the adaptive waveform design is fused with state of the art adaptive receiver techniques to conceive a novel adaptive MIMO radar system under practical constraints in this thesis