Array filters for attenuation of interference arriving from a wide range of angles

The material presented in this thesis, deals with the application of digital signal processing techniques for filtering and retrieval of a desired signal from a one-dimensional array of sensor recorded signals. The model and equation used for implementation takes into account a desired signal mixed with an interfering signal, plus a random noise factor. The interference could be from a number of different sources. It could be a reflected source signal (as in the case of siesmic study) or an interfering signal generated from a separate source for which amplitudes of up to five times that of the source signal are used. The design of the array filter takes into account an uncertainty about the direction of the interference signal. An uncertainty window is used to indicate the highest probability of the interference arrival angle. The coherent interference is expected to be attenuated better inside the uncertainty region but since there exists an all pass condition for the desired signal, some relation between the seperation angle between the desired and the interference signal is expected. There are several examples of the filter output presented here through the processing of synthetic signals. The synthetic signals were generated through a computer program representing the Bernoulli-Gaussian random sequences. The examples presented weigh trade offs between variation of filter, arrival angle, and uncertainty region parameters. The filter was implemented in Fortran-77 using the IMSL routines, and Linpack a linear equation solution software package