LoboVault Home

Optimal digital filter design for dispersed signal equalization

LoboVault

Please use this identifier to cite or link to this item: http://hdl.handle.net/1928/3286

Optimal digital filter design for dispersed signal equalization

Show full item record

Title: Optimal digital filter design for dispersed signal equalization
Author: Mihalik, Andrew
Advisor(s): Pattichis, Marios
Committee Member(s): Doerry, Armin
Christodoulou, Christos
Department: University of New Mexico. Dept. of Electrical and Computer Engineering
Subject: Optimal
Equalization
Dispersion
Digital
Ionosphere
Filter
Optimization
Dispersed
Trust Region
Satellite
refraction
allpass
biquad
bi-quad
dispersive
group delay
pole
zero
pulse
frequency
cascade
matched
LC Subject(s): Signal processing
Degree Level: Masters
Abstract: Any signal a satellite receives from Earth has traveled through the ionosphere. Transmission through the ionosphere results in a frequency dependent time-delay of the signal frequency components. This effect of the medium on the signal is termed dispersion, and it increases the difficulty of pulse detection. A system capable of compensating for the dispersion would be desirable, as pulsed signals would be more readily detected after compression. In this thesis, we investigate the derivation of a digital filter to compensate for the dispersion caused by the ionosphere. A transfer function model for the analysis of the ionosphere as a system is introduced. Based on the signal model, a matched filter response is derived. The problem is formulated as a group delay compensation effort. The Abel-Smith algorithm is employed for the synthesis of a cascaded allpass filter bank with desired group delay characteristics. Extending this work, an optimized allpass filter is then derived using a pole location approach. A mean-square error metric shows that the optimized filter can reproduce, and even improve upon, the results of the Abel-Smith design with a significantly lower order filter. When compared against digital filters produced with the least p-th minimax algorithm, we find that the new method exhibits significantly lower error in the band of interest, as well as lower mean squared error overall. The result is a simple optimized equalization filter that is stable, robust against cascading difficulties, and applicable to arbitrary waveforms. This filter is the cornerstone to a new all-digital electromagnetic pulse detection system.
Graduation Date: July 2007
URI: http://hdl.handle.net/1928/3286


Files in this item

Files Size Format View
Optimal_Digital ... ersed_Signal_Equalizat.pdf 1.088Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record

UNM Libraries

Search LoboVault


Browse

My Account