Applications of Wavelet Transform and Artificial Neural Network in Digital Signal Detection for Indoor Optical Wireless Communication Sujan Rajbhandari.

Slides:



Advertisements
Similar presentations
Z. Ghassemlooy, S Rajbhandari and M Angelova
Advertisements

Chapter : Digital Modulation 4.2 : Digital Transmission
Applications in Signal and Image Processing
Introduction and Overview Dr Mohamed A. El-Gebeily Department of Mathematical Sciences KFUPM
S. Rajbhandari, Prof. Z. Ghassemlooy, Prof. M. Angelova School of Computing, Engineering & Information Sciences, University of Northumbria, Newcastle upon.
CHAPTER 4 DIGITAL MODULATION Part 1.
Machine Learning Neural Networks
ICTON 2007, Rome, Italy The Performance of PPM using Neural Network and Symbol Decoding for Diffused Indoor Optical Wireless Links 1 S. Rajbhandari, Z.
0 - 1 © 2007 Texas Instruments Inc, Content developed in partnership with Tel-Aviv University From MATLAB ® and Simulink ® to Real Time with TI DSPs Wavelet.
Contact: Neural Networks for PRML equalisation and data detection What is Partial Response signalling ? Some commonly used PR.
Sujan Rajbhandari PGNET Performance of Convolutional Coded Dual Header Pulse Interval Modulation in Infrared Links S. Rajbhandari, Z. Ghassemlooy,
RAKE Receiver Marcel Bautista February 12, Propagation of Tx Signal.
Z. Ghassemlooy & S Rajbhandari
Wavelet Transform A very brief look.
Communication Systems
Wavelet Transform. What Are Wavelets? In general, a family of representations using: hierarchical (nested) basis functions finite (“compact”) support.
Multi-Resolution Analysis (MRA)
Signal Analysis and Processing for SmartPET D. Scraggs, A. Boston, H Boston, R Cooper, A Mather, G Turk University of Liverpool C. Hall, I. Lazarus Daresbury.
Introduction to Wavelets
Wavelet-based Coding And its application in JPEG2000 Monia Ghobadi CSC561 project
IASTED- WOC- Canada 07 1 CONVOLUTIONAL CODED DPIM FOR INDOOR NON-DIFFUSE OPTICAL WIRELESS LINK S. Rajbhandari, Z. Ghassemlooy, N. M. Adibbiat, M. Amiri.
ECE 4730: Lecture #10 1 MRC Parameters  How do we characterize a time-varying MRC?  Statistical analyses must be used  Four Key Characteristics of a.
Introduction to Wavelets -part 2
ECE 501 Introduction to BME ECE 501 Dr. Hang. Part V Biomedical Signal Processing Introduction to Wavelet Transform ECE 501 Dr. Hang.
Communication Systems
ENG4BF3 Medical Image Processing
Formatting and Baseband Modulation
Convolutional Neural Networks for Image Processing with Applications in Mobile Robotics By, Sruthi Moola.
Presented by: Kamakhaya Argulewar Guided by: Prof. Shweta V. Jain
CE 4228 Data Communications and Networking
The Wavelet Tutorial: Part3 The Discrete Wavelet Transform
Dept. of EE, NDHU 1 Chapter Three Baseband Demodulation/Detection.
1 Techniques to control noise and fading l Noise and fading are the primary sources of distortion in communication channels l Techniques to reduce noise.
1 Wavelets, Ridgelets, and Curvelets for Poisson Noise Removal 國立交通大學電子研究所 張瑞男
EELE 5490, Fall, 2009 Wireless Communications Ali S. Afana Department of Electrical Engineering Class 5 Dec. 4 th, 2009.
Performance of Discrete Wavelet Transform – Artificial Neural Network Based Signal Detector/Equalizer for Digital Pulse Interval Modulation in Practical.
WAVELET (Article Presentation) by : Tilottama Goswami Sources:
COMPARISON OF IMAGE ANALYSIS FOR THAI HANDWRITTEN CHARACTER RECOGNITION Olarik Surinta, chatklaw Jareanpon Department of Management Information System.
Wavelet-based Coding And its application in JPEG2000 Monia Ghobadi CSC561 final project
EE 6331, Spring, 2009 Advanced Telecommunication Zhu Han Department of Electrical and Computer Engineering Class 7 Feb. 10 th, 2009.
The Physical Layer Lowest layer in Network Hierarchy. Physical transmission of data. –Various flavors Copper wire, fiber optic, etc... –Physical limits.
Artificial Neural Networks Bruno Angeles McGill University – Schulich School of Music MUMT-621 Fall 2009.
Handwritten Recognition with Neural Network Chatklaw Jareanpon, Olarik Surinta Mahasarakham University.
Wavelets and Multiresolution Processing (Wavelet Transforms)
Wavelets Anderson G Moura 05/29/2015. Introduction Biomedical signals usually consist of brief high-frequency components closely spaced in time, accompanied.
1 Wavelet Transform. 2 Definition of The Continuous Wavelet Transform CWT The continuous-time wavelet transform (CWT) of f(x) with respect to a wavelet.
COMMUNICATION SYSTEM EEEB453 Chapter 5 (Part IV Additional) DIGITAL TRANSMISSION.
When a signal is transmitted over a channel, the frequency band and bandwidth of the channel must match the signal frequency characteristics. Usually,
Time frequency localization M-bank filters are used to partition a signal into different frequency channels, with which energy compact regions in the frequency.
Neural Networks Presented by M. Abbasi Course lecturer: Dr.Tohidkhah.
Supervised learning network G.Anuradha. Learning objectives The basic networks in supervised learning Perceptron networks better than Hebb rule Single.
Spectrum Sensing In Cognitive Radio Networks
WAVELET AND IDENTIFICATION WAVELET AND IDENTIFICATION Hamed Kashani.
APPLICATION OF A WAVELET-BASED RECEIVER FOR THE COHERENT DETECTION OF FSK SIGNALS Dr. Robert Barsanti, Charles Lehman SSST March 2008, University of New.
Sujan Rajbhandari LCS Convolutional Coded DPIM for Indoor Optical Wireless Links S. Rajbhandari, N. M. Aldibbiat and Z. Ghassemlooy Optical Communications.
Face Detection Using Neural Network By Kamaljeet Verma ( ) Akshay Ukey ( )
By Dr. Rajeev Srivastava CSE, IIT(BHU)
Wavelet Transforms ( WT ) -Introduction and Applications
Presenter : r 余芝融 1 EE lab.530. Overview  Introduction to image compression  Wavelet transform concepts  Subband Coding  Haar Wavelet  Embedded.
WAVELET NOISE REMOVAL FROM BASEBAND DIGITAL SIGNALS IN BANDLIMITED CHANNELS Dr. Robert Barsanti SSST March 2010, University of Texas At Tyler.
Deep Learning Overview Sources: workshop-tutorial-final.pdf
By Poornima Balakrishna Rajesh Ganesan George Mason University A Comparison of Classical Wavelet with Diffusion Wavelets.
An Experimental Receiver Design For Diffuse IR Channels Based on Wavelet Analysis & Artificial Intelligence R J Dickenson and Z Ghassemlooy O ptical C.
Wavelet domain image denoising via support vector regression
PERFORMANCE OF A WAVELET-BASED RECEIVER FOR BPSK AND QPSK SIGNALS IN ADDITIVE WHITE GAUSSIAN NOISE CHANNELS Dr. Robert Barsanti, Timothy Smith, Robert.
Techniques to control noise and fading
ARTIFICIAL NEURAL NETWORKS
H7 By: Myron E. Hohil Sachi Desai
Wavelet Transform Fourier Transform Wavelet Transform
Wavelet transform application – edge detection
Presentation transcript:

Applications of Wavelet Transform and Artificial Neural Network in Digital Signal Detection for Indoor Optical Wireless Communication Sujan Rajbhandari 1 Supervisors Prof. Maia Angelova Prof. Z. Ghassemlooy Prof. Jean-Pierre Gazeau

Optical Wireless Communication Sujan Rajbhandari 2  Light as the carrier of information  Also popularly known as free space optics (FSO) or Free Space Photonics (FSP) or open-air photonics.  Indoor or outdoor

Transmission Format Transmitted signal  ‘1’ presence of an optical pulse  ‘0’ absence of an optical pulse Sujan Rajbhandari

Links Sujan Rajbhandari 4 Non-LOS  Multipath Propagation  Intersymbol interference (ISI)  Difficult to achieve high data rate if ISI is not mitigated. Non-LOS  Multipath Propagation  Intersymbol interference (ISI)  Difficult to achieve high data rate if ISI is not mitigated. Rx Tx  LOS LOS  No multipath propagation  Noise and device speed are limiting factors  Possibility of blocking LOS  No multipath propagation  Noise and device speed are limiting factors  Possibility of blocking Tx Rx

Received Signal Sujan Rajbhandari 5 Non-LOS LOS

Classical Digital Signal Detection  Set a threshold level.  Compared the received signal with the threshold level  Set ‘1’ if received signal is greater than threshold level  Set ‘0’ is received signal is less than threshold level. Sujan Rajbhandari 6

Classical signal detection techniques: Assumptions  The statistical of noise is known.  Maximise the signal to noise ratio for unknown noise with known statistics.  Channel characteristics are known( at least partially ) and generally assume to be linear.

Digital signal Reception: Problem of feature extraction and pattern classification 8 Received signal  ‘1’ signal + interference  ‘0’ interference only (noise and intersymbol interference (ISI)). Interference only signal + interference Sujan Rajbhandari

Receiver from the Viewpoint of Statistics 9  Testing a Null Hypothesis of a)Received signal is interference only against b)Alternative Hypothesis of received signal is signal plus interference Sujan Rajbhandari

Problem of Feature Extraction and Pattern Classification 10  Receiver Block diagram Optical Receiver Wavelet Transform Artificial Neural Network Threshold Detector Feature Extraction Feature Extraction Pattern Classification Pattern Classification Sujan Rajbhandari

Time- Frequency analysis Fourier Transform  Time-frequency mapping  What frequencies are present in a signal but fails to give picture of where those frequencies occur.  No time resolution. Sujan Rajbhandari 11

Time- Frequency analysis Windowed Fourier Transform (Short time Fourier transform)  Chop signal into equal sections  Find Fourier transform of each section Disadvantages  Problem how to cut a signal  Fixed time and frequency resolution Sujan Rajbhandari 12

Time- Frequency analysis Continuous Wavelet Transform (CWT)  Vary the window size to vary resolution (Scaling).  Large window for precise low-frequency information, and shorter window high-frequency information  Based on Mother wavelet.  Mother Wavelet are well localised in time.(Sinusoidal wave which are the based of Fourier transform extend from minus infinity to plus infinity) Sujan Rajbhandari 13

Continues Wavelet Transform  Where are wavelets and s and τ are scale and translation.  Translation time resolution  scale frequency resolution  Wavelets are generated from scaling and translation the Mother wavelet. CWT of Signal f(t) and reconstruction is given by

Discrete Wavelet Transform Dyadic scales and positions DWT coefficient can efficiently be obtained by filtering and down sampling 1 1 Mallat, S. (1989), "A theory for multiresolution signal decomposition: the wavelet representation," IEEE Pattern Anal. and Machine Intell., vol. 11, no. 7, pp

Artificial Neural Network  Fundamental unit : a neuron  Based on biological neuron  Capability to learn Sujan Rajbhandari 16 b wnwn y x1x1 f(.)f(.) ∑ w1w1 Output xnxn...

Artificial Neural Network  Input layer, hidden layer(s) and output layer  Extensively used as a classifier  Supervised and unsupervised learning.  Weight are adjust by comparing actual output and target output Sujan Rajbhandari 17

Feature Extraction: Discrete Wavelet Transform Sujan Rajbhandari 18 DWT of Interference only DWT of signal +Interference Significant difference in approximation coefficient,a 3. No difference in other details coefficients. (Details coefficient are due to the high frequency component of signal, mainly due to noise.) Significant difference in approximation coefficient,a 3. No difference in other details coefficients. (Details coefficient are due to the high frequency component of signal, mainly due to noise.)

Denoising  The high frequency component can be removed or suppressed.  Two Approach Taken 1. Threshold approach in which the detail coefficients are suppressed by either ‘hard’ or ‘soft’ thresholding. 2. Coefficient removal approach in which detail coefficients are completely removed by making it zero. Sujan Rajbhandari 19

De-noised Signal Sujan Rajbhandari 20 LOS Links Non-LOS Links Denoising effectively removes high frequency component. Equalization is necessary for non-LOS links Identical performance for both de-noising approaches.

21 Artificial Neural Network : Pattern Classifier  Artificial Neural Network can be trained to differentiate the interference from signal plus interference.  DWT are fed to ANN.  ANN is first trained to classify by providing examples.  ANN can be utilized both as a pattern classifier and equalizer.

Results Sujan Rajbhandari 22  The Coefficient removal approach (CRA) of denoising gives the best result.  Easier to train ANN using CRA as the DWT coefficients are removed by 8 folds if 3 level of DWT is taken.  Effective for detection and equalization. Figure: The Performance of On-off Keying at 150Mbps for diffused channel with a D rms of 10ns

Comparison with traditional methods Maximum performance of 8.6dBcompared to linear equalizer performance depends on the mother wavelets. Discrete Meyer gives the best performance and Haar the worst performance among studied mother wavelet

Conclusion  Digital signal detection can be reformulated as feature extraction and pattern classification.  Discrete wavelet transform is used for feature extraction.  Artificial Neural Network is trained for pattern classification.  Performance can further be enhance by denoising the signal before classifying it. Sujan Rajbhandari 24

25 Thank You Discussions