東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Neural network detection for satelli...

Yuan, Jun.

Linked to FindBook

Google Book

Amazon

博客來

Neural network detection for satellite mobile communications.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Neural network detection for satellite mobile communications./
Author:	Yuan, Jun.
Description:	108 p.
Notes:	Source: Masters Abstracts International, Volume: 42-04, page: 1350.
Contained By:	Masters Abstracts International42-04.
Subject:	Engineering, Electronics and Electrical. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=MQ86204
ISBN:	0612862046

Neural network detection for satellite mobile communications.
Yuan, Jun.

Neural network detection for satellite mobile communications. - 108 p.

Source: Masters Abstracts International, Volume: 42-04, page: 1350.

Thesis (M.Sc.(Eng))--Queen's University at Kingston (Canada), 2003.

In the first part of this thesis, we derive generalized formulas to calculate the exact average symbol error rate (SER) of satellite mobile channels. These formulas can be used to evaluate both the effects of high-power amplifier (HPA) nonlinearity and flat fading. In the second part, we propose two adaptive neural network receivers that can be used in conjunction with well-known techniques, like adaptive filters or pilot-symbol-aided method, to overcome the problems of nonlinearity and fading. The fast neural network learning algorithm (i.e. natural gradient descent) is employed, which outperforms the ordinary gradient descent (i.e. back propagation) algorithm in terms of convergence speed and modeling accuracy. Since the neural network is used as a key technology in receiver design to overcome the nonlinear problem caused by HPA, it is important for system designers to understand its learning behavior and performance capabilities. The last part of the thesis, therefore, investigates a statistical analysis of natural gradient neural network learning in the case of finite input elements. (Abstract shortened by UMI.)

ISBN: 0612862046Subjects--Topical Terms:

626636
Engineering, Electronics and Electrical.

Neural network detection for satellite mobile communications.
LDR:02009nmm 2200277 4500 001 1848277
005 20051102155248.5
008 130614s2003 eng d
020 $a 0612862046
035 $a (UnM)AAIMQ86204
035 $a AAIMQ86204
040 $a UnM $c UnM
100 1 $a Yuan, Jun. $3 1044729
245 1 0 $a Neural network detection for satellite mobile communications.
300 $a 108 p.
500 $a Source: Masters Abstracts International, Volume: 42-04, page: 1350.
500 $a Adviser: Mohamed Ibukahla.
502 $a Thesis (M.Sc.(Eng))--Queen's University at Kingston (Canada), 2003.
520 $a In the first part of this thesis, we derive generalized formulas to calculate the exact average symbol error rate (SER) of satellite mobile channels. These formulas can be used to evaluate both the effects of high-power amplifier (HPA) nonlinearity and flat fading. In the second part, we propose two adaptive neural network receivers that can be used in conjunction with well-known techniques, like adaptive filters or pilot-symbol-aided method, to overcome the problems of nonlinearity and fading. The fast neural network learning algorithm (i.e. natural gradient descent) is employed, which outperforms the ordinary gradient descent (i.e. back propagation) algorithm in terms of convergence speed and modeling accuracy. Since the neural network is used as a key technology in receiver design to overcome the nonlinear problem caused by HPA, it is important for system designers to understand its learning behavior and performance capabilities. The last part of the thesis, therefore, investigates a statistical analysis of natural gradient neural network learning in the case of finite input elements. (Abstract shortened by UMI.)
590 $a School code: 0283.
650 4 $a Engineering, Electronics and Electrical. $3 626636
650 4 $a Artificial Intelligence. $3 769149
690 $a 0544
690 $a 0800
710 2 0 $a Queen's University at Kingston (Canada). $3 1250078
773 0 $t Masters Abstracts International $g 42-04.
790 1 0 $a Ibukahla, Mohamed, $e advisor
790 $a 0283
791 $a M.Sc.(Eng)
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=MQ86204