Abstract:
The research presented in this thesis is about a task of geolocation of radio frequency
emitters. In this research the problem of geolocation of non-collaborative emitter was
addressed. This thesis presents the novel algorithm for the RF emitter geolocation
based on the image process technique known as Hough Transform. The comparison
of this algorithm with traditional approaches to geolocation showed a number of
benefits, like robustness, accuracy and advanced fusion capability. The application
of the Hough Transform to data fusion allowed to use the modern concepts of agentbased
fusion and cluster level fusion, thus moving the solution of the problem of the
geolocation to upper level of fusion hierarchy. The work on Hough Transform lead
to a comparison of the Bayesian and non-Bayesian approaches in solving the task of
geolocation. Exploitation of the comparison lead to the derivation of a generalized
estimator. This estimator highlighted a number of mathematical functions which
can be exploited for geolocation and data fusion. These functions has been tested
for the purpose of data fusion in geolocation and it was found that Hough Transform
is a useful alternative approach for the data fusion for geolocation of RF emitter.
