Signal Processing Application

This experiment was performed in group of five. Our group topic was Speech Recognition.

Group Members: Abhisu Mishra, Nikhil Nandoskar, Meet Nathvani, Aditya Parkhi, Himanshu  Pathak.

Problem Definition:1)Acquisition and display of Raw bioelectrical signal.

                             2)Filtering and Conditioning of raw signal.

                             3)Plotting of processed bio-electric signal. 

Paper Review:

Published On: 10th October,2012

Author : A Roman Gonzalez.

Summary : The work presented in this paper is based on Roman Gonzalez.This paper provides the basis and foundation for developing a brain computer interface BCI and various stages of processing and analyzing the different techniques currently used.

Patent Review:

Patent no: EP2204929

Date of Patent: 7/7/2010

Inventor : Lanzafame Pietro, Zona Paolo, Galli, Giovanni Bramanti Placido.

Summary : Signal acquiston and conditioning system for bioelectric signal compramises an acquistion and pre conditioning device containing a DSP processor which accepts input bioelectric signal in a different manner with hardly any other circuitry required.

Links:

Paper:https://hal.inria.fr/file/index/docid/742211/filename/EEG_Signal_Processing_for_BCI_Applications_.pdf

Patent: https://www.google.com/patents/EP2204929A1?cl=en

Basic Operations using DSP Processor

This was a demonstration experiment where one of our seniors showed us the operations that can be done on a dsp processor. The various operations carried out were Arithmetic Operations which included Addition Subtraction Multiply Divide, Logical Operations like And and Not and Shifting Operations like Logical Shift LEft, Logical Shitft Right, Rotate Right, Rotate Left.

Digital FIR Filter Design using Frequency Sampling Method

In this experiment, we were required to design a digital FIR filter using Frequency Sampling Method (FSM). Scilab was used to implement it.Here we observed that phase spectrum is linear within the positive lobs of magnitude spectrum. Discontinuity is observed i phase plot between lobes and also when the phase spectrum goes out of the range of -pi to +pi.

https://drive.google.com/file/d/0B9zlXLFfOipjQUZ3Q3pvU1RRTVUyMWpFdUJ4djd6UzltVlU4/view?usp=sharing

Digital FIR Filter Design using Windowing Method

The user had to input values like Attenuation in Stop band (As) and Pass band (Ap) as well as Pass band frequency, Stop band frequency and sampling frequency.The window function is required to truncate infinite samples of hd(n). A high pass filter and a bandpass filter was designed.The phase  plot being linear there will be no distortion at the output.

https://drive.google.com/file/d/0B9zlXLFfOipjMllRandkZG5tamk1NFEwWWsxQkJHMlYyT0Y0/view?usp=sharing

Digital Chebyshev Filter Design

 Low pass and High Pass filters were designed with the input parameters As ,Ap pass band frequency and stop band frequency and sampling frequency.The order of the filter was calculated using the given values of attenuation in pass band and stop band, pass band and stop band frequencies and sampling frequency. The normalised and denormalised H(s) was calculated, from which the Transfer Function H(z) was calculated. For HPF there is a pole at +1 and for LPF pole is at -1.

https://drive.google.com/file/d/0B9zlXLFfOipjdzZMa240RU95VG1TUl9xODJVOGtyaHFoME9J/view?usp=sharing

Digital Butterworth Filter Design

Low pass and High Pass filters were designed with the input parameters As ,Ap pass band frequency and stop band frequency and samppling frequency.The pole zero plot was also drawn and it was observed that all the poles lie inside the unit circle indication both the High Pass and Low Pass fiters were stable.It was observed that  the butterworth filter is monotonic in it's pass band as well as stop band meaning it does not have any ripples.

https://drive.google.com/file/d/0B9zlXLFfOipjd1Q5UEkyaWtwQnAwVXZyU1l0V0JWRFZ5WGNr/view?usp=sharing

Overlap Add Method / Overlap Save Method

Linear convolution was performed first by using OAM method and then OSM Method. There were 4 decomposed signals and N=8. OAM and OSM are efficient ways of calculating very long signal x[n] and a finite impulse response h[n].

https://drive.google.com/file/d/0B9zlXLFfOipjbWhkdVFhb0g1TEk/view?usp=sharing

https://drive.google.com/file/d/0B9zlXLFfOipjQ05IZS1wcTFjX1E/view?usp=sharing

Fast Fourier Transform

4 point FFT was performed and the output was stored in X[k]. DFT and IDFT were compared on the points like Complex Multiplication, Complex Addition and Real Addition and Real Multiplications. Number of computations required are less in FFT and hence the speed is increased.

https://drive.google.com/file/d/0B9zlXLFfOipjdzRJX0J2ZjBxeVE/view?usp=sharing

https://drive.google.com/file/d/0B9zlXLFfOipjRl9KdlY0V05YV2c/view?usp=sharing

Convolution and Correlation Algorithms

This experiment involved all the topic we had studied in Sem 5 Signals and Systems

They were done as follows:

1. Linear Convolution

L and M values were taken with L=5 and M=4. The output was stored in y(n) and the length of the op signal was calculated in the program as N=LM-1.

2.Circular Convolution

It was implemented using different values of N. Aliasing effect was observed in circular convolution.

Also Linear using circular convolution was done for L=4 and M=3. The length of the op signal was calculated as N=L+M-1.

3. Correlation

Correlation was studied with L=4 and M=3. The length of the op signal was calculated as N=L+M-1.

The output was in the form of palindrome when the signals were same. It can be used to find error in the signal. The op of correlation as both sided.

https://drive.google.com/file/d/0B9zlXLFfOipjc3Q5QXdrMzVxQk0/view?usp=sharing
https://drive.google.com/file/d/0B9zlXLFfOipjem5abGNOTG9KY2s/view?usp=sharing
https://drive.google.com/file/d/0B9zlXLFfOipjTFRjUWFFWVAzVzA/view?usp=sharing
https://drive.google.com/file/d/0B9zlXLFfOipjR0JmN3FkQzBwclk/view?usp=sharing

Discrete Fourier Transform

In this eperiment we took entries of x(n) and found out its Discrete Fourier Transform. 
It was done for two values of N namely N=4 and N=8. IDFT was performed to verify the result.We observed that as the spacing between the values reduces the value of N increases. Also the quality of spectrum obtained improves. By apending the input signal by zeros the resolution error reduces. 
https://drive.google.com/file/d/0B9zlXLFfOipjbkloRXgyQVMtYWM/view?usp=sharing 
https://drive.google.com/file/d/0B9zlXLFfOipjTEhkY2xpSlpfUjQ/view?usp=sharinghttps://drive.google.com/file/d/0B9zlXLFfOipjTEhkY2xpSlpfUjQ/view?usp=sharing