Channel Estimation and Modeling for Wireless Underwater Acoustic Communications
Department
Electrical and Computer Engineering
Major
Electrical Engineering
Research Advisor
Zheng, Y. Rosa
Advisor's Department
Electrical and Computer Engineering
Funding Source
National Science Foundation Grant ECCS-0846486; Office of Naval Research Grant N00014-07-1-0219
Abstract
Many channel estimation methods are based upon stochastic models. It has been well established that the envelope of the radio channel impulse response (CIR) probability distribution function is a Rayleigh distribution. This results from the Gaussian distributed real and imaginary components. However, the underwater acoustic CIR has not been well established. Since the underwater acoustic channel is more severe than radio frequency channels, a t location-scale based, rather than Gaussian based, model for the underwater acoustic CIR complex components is proposed. An acoustic channel impulse response is estimated from existing experimental data. Then, a t location-scale distribution is fitted to the Probability Distribution Function of the estimated channel impulse response. Lastly, the fitted t location-scale distribution is compared to a fitted Gaussian distribution. It will be shown the t location-scale distribution is a better choice than the Gaussian distribution by using the two sample Kolmogorov-Smirnov test.
Biography
Jesse Cross is a senior in the Electrical and Computer Engineering department. He is expecting to graduate in May of 2010.
Research Category
Engineering
Presentation Type
Oral Presentation
Document Type
Presentation
Location
Gasconade Room
Presentation Date
07 Apr 2010, 9:00 am - 9:30 am
Channel Estimation and Modeling for Wireless Underwater Acoustic Communications
Gasconade Room
Many channel estimation methods are based upon stochastic models. It has been well established that the envelope of the radio channel impulse response (CIR) probability distribution function is a Rayleigh distribution. This results from the Gaussian distributed real and imaginary components. However, the underwater acoustic CIR has not been well established. Since the underwater acoustic channel is more severe than radio frequency channels, a t location-scale based, rather than Gaussian based, model for the underwater acoustic CIR complex components is proposed. An acoustic channel impulse response is estimated from existing experimental data. Then, a t location-scale distribution is fitted to the Probability Distribution Function of the estimated channel impulse response. Lastly, the fitted t location-scale distribution is compared to a fitted Gaussian distribution. It will be shown the t location-scale distribution is a better choice than the Gaussian distribution by using the two sample Kolmogorov-Smirnov test.
