Adaptive algorithms for a GPS interference suppression receiver and a sparse reconfigurable adaptive filter

GPS is a one-way satellite-based navigation system employing spread-spectrum techniques that is widely used for commercial and military applications. Although the very low signal-to-noise ratio is handled by the large spreading gain, GPS is susceptible to high-power interference signals and various types of jammers. In this dissertation, we propose multicomponent receiver architectures for GPS interference suppression. A conventional antenna system is first considered which utilizes a minimum-variance distortionless-response beamformer and assumes that the GPS signal angle of arrival (AOA) and the antenna model are known at the receiver. The sensitivity problem of this system is eliminated by a multicomponent system based on a multistage matched filter (MF). Since this MF receiver also has a high computational complexity because the jammer AOAs must be estimated, we introduce a blind interference canceler based on the constant modulus (CM) array with a low computational complexity.; In order to make the system more efficient, we propose two new despreaders: a null despreader and a modified despreader. A blind adaptive GPS receiver based on a null despreader consists of a conventional GPS despreader and a so-called null despreader which together modify the received signal so that the CM array can extract the GPS signal of interest. The modified despreading receiver is a lowcomplexity blind adaptive receiver that is based on a novel modified despreader and the CM array. We also consider the multiple satellite problem and extend the proposed receiver to capture several CPS signals of interest.; This dissertation also considers an application of the adaptive algorithm for a reconfigurable photonic switch. A nonblocking photonic switch can be used to implement a tapped delay line with a large number of adaptive weights and a wide range of time delays. In this dissertation, we present a sparse reconfigurable adaptive filter (SRAF) based on a photonic switch with an input/output connection architecture that can be represented by a matrix of adaptive weights. For this filter, we propose an adaptive algorithm based on a system identification formulation that adapts the weights and chooses the appropriate delay combinations, which has good performance for white and non-white input signals.