Simultaneous Signal Separation and Denoising in DCT Domain for Wireless Communication Applications

F. E. ABD, EL-SAMIE and M., ELKORDY and O., ZAHRAN and R., ALI (2015) Simultaneous Signal Separation and Denoising in DCT Domain for Wireless Communication Applications. In: Third International Conference on Advances in Computing, Communication and Information Technology- CCIT 2015, 26 - 27 May,2015, Birmingham B42 2SU, UNITED KINGDOM.

[img]
Preview
Text
20150604_090101.pdf - Published Version

Download (902kB) | Preview
Official URL: https://www.seekdl.org/conferences/paper/details/6...

Abstract

This paper addresses the problem of blind signal separation (BSS) for the system of multiple input and multiple output signals (MIMO). Different modulation techniques such as quadrature phase shift keying (QPSK), minimum shift keying (MSK), and Gaussian minimum shift keying (GMSK) have been considered. Several methods have been investigated to solve this problem such as principle component analysis (PCA), independent component analysis (ICA), and multi user kurtosis (MUK) algorithms. Different modulation techniques and different algorithms in the separation have been investigated to for comparison between results. In this paper, we propose wavelet denoising with PCA, ICA and MUK methods. We use the separation algorithm for the discrete cosine transforms (DCT) for blind of mixed signals, instead of separating the mixtures themselves, as a technique that achieves a great result in eliminating the noise. The proposed algorithm is used with multi user kurtosis (MUK) that used for blind of mixed signals with considering the instantaneous mixture of two sources. The simulation results show a considerable improvement in extracted signals when compared to original signals.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: BSS, MIMO, PCA, ICA, MUK, MSK, GMSK, QPSK, DCT MSE, SNR
Depositing User: Mr. John Steve
Date Deposited: 30 Apr 2019 06:00
Last Modified: 30 Apr 2019 06:00
URI: http://publications.theired.org/id/eprint/1699

Actions (login required)

View Item View Item