Characterization of antenna radiation and receiving properties in complex environments based on physical models
MetadataShow full item record
In this dissertation, physics-based signal processing techniques are applied to various aspects of antenna radiation and receiving problems in the presence of complex environments. The overall objective is to characterize antenna properties more effectively and relate them directly to features of the environment via physical models. According to the interaction distance, the problem is studied at four different scales: the induced current, the antenna element, the mounting platform and the propagation channel. Different models are applied at each scale to characterize the underlying physics. First, a frequency-dependent time-of-arrival model is proposed to represent the frequency response of the induced current. The model parameters are extracted at a few computations at low frequencies and the broadband response of the current is then extrapolated with the model. Second, the mutual coupling effect in an antenna array is studied and the standard coupling matrix model is reviewed and analyzed. The applicable condition of the model is derived. The model is further extended to characterize the coupling in more complicated wire antenna arrays. Third, a point radiator model is applied to represent the radiation pattern of antennaplatform radiation. A matching pursuit algorithm is proposed to extract the model parameters. The radiation pattern of an antenna-platform system can be sparsely represented and efficiently reconstructed from the model. This model is further combined with the mutual coupling model to characterize the array responses with mounting platforms. Finally, two statistical models are derived to model the propagation channel in a complex environment. The slow fade and the direction-ofarrival statistics are characterized in a multipath channel based on measurement data in an urban environment. Simulation of the propagation channel based on the models is shown to have very similar statistical properties to the measurement.