TRACKING ANALYSIS OF C -NLMS AND LEAKY C-NLMS ALGORITHMS FOR COLORED GAUSSIAN INPUTS

TRACKING ANALYSIS OF C -NLMS AND LEAKY C-NLMS ALGORITHMS FOR COLORED GAUSSIAN INPUTS. Masters thesis, King Fahd University of Petroleum and Minerals.

[img]
Preview
PDF (Master Thesis)
Thesis.pdf

Download (783kB) | Preview

Arabic Abstract

احدى اهم المميزات الخاصة بالمنقحات المتكيفة هو قدرتها على عمل تتبع للتغيرات الحاصلة في خصائص الاشارة الكهربائية. وهذا كله يرجع الى حقيقة ان هذه المنقحات تعتمد على المعلومات اللحظية, الخصائص الاحصائية لاشارة الخطأ بالاضافة الى متجه الوزن والتي بدورها تقوم بالاستجابة لهذه التغيرات الحاصلة في خصائص الاشارة الكهربائية.

English Abstract

One of the main features of the adaptive filters is its ability to track the variations in the underlaying signal statistics. This is due to the fact that they are relying on the instantaneous data, the statistical properties of the error signal and the weight vector of the adaptive lter will react to these variations in the input signal properties. In this thesis, the tracking analysis of C-NLMS algorithm and leaky C-NLMS algorithm are investigated in a nonstationary environments, the environment is characterized by a rst order autoregressive model and the lter input is assumed to be correlated gaussian, A closed form EMSE expression is derived for both algorithms by relying on energy conservation and weighted energy conservation relations, Finally, simulation results show a great match between the analytical and the simulation results especially in the slow changable environments.

Item Type: Thesis (Masters)
Subjects: Research > Information Technology
Electrical
Department: College of Engineering and Physics > Electrical Engineering
Committee Advisor: Zerguine, Azzedine
Committee Members: Abdul-Jauwad, S. H. and Zidouri, Abdelmalek and Abu-Al-Saud, Wajih
Depositing User: ANKAR (g200603940) (g200603940)
Date Deposited: 31 May 2011 12:09
Last Modified: 01 Nov 2019 15:29
URI: http://eprints.kfupm.edu.sa/id/eprint/136491