Font Size: a A A

Wireless channel state and model parameter estimation

Posted on:2007-09-14Degree:Ph.DType:Thesis
University:University of Alberta (Canada)Candidate:Chen, YunfeiFull Text:PDF
GTID:2458390005488385Subject:Engineering
Abstract/Summary:
Practical wireless communication channels are usually characterized by models. The models are determined by parameters. They are also represented by realizations in specific observation intervals that determine channel state. This thesis contributes to the estimation of both channel state and model parameters.;Second, maximum likelihood and moment-based estimators for the channel model parameters are proposed by using noisy channel samples. The estimators operate with or without knowledge of the noise power. Also, maximum likelihood estimators for the Ricean K parameter are derived by using fading phase samples, a method not considered previously.;Finally, maximum likelihood estimation of signal-to-noise ratio is studied. Two measures of signal-to-noise ratio are considered. The performances of the estimators are analyzed under the assumption of no decision errors. Using both known and unknown symbols in a frame, an approximate maximum likelihood estimator for signal-to-noise ratio is derived. A non-data-aided moment-based estimator for signal-to-interference-plus-noise ratio in a quadrature amplitude modulation system is also developed. In other works, maximum likelihood estimation of the average signal-to-noise ratio and a joint estimation of the K parameter and the average signal-to-noise ratio in a Ricean fading channel are performed. As the last part of this thesis, some concluding remarks are made and future possible works are outlined.;First, maximum likelihood decision-based estimators for the channel state parameters are developed. Effects of channel estimation errors on the performances of two selection diversity combiners are evaluated. Novel diversity receivers using statistics of the channel estimation errors are designed. Optimum pilot symbol assisted modulation using pilot symbols for channel state parameter estimation is also investigated. As well, novel non-data-aided maximum likelihood estimators for the channel state parameters in an ultra-wide bandwidth system are derived, and the Cramer-Rao lower bounds are calculated analytically.
Keywords/Search Tags:Channel, Parameter, Estimation, Model, Maximum likelihood, Signal-to-noise ratio
Related items