Experiment 10 - DSP Application.

This was a group experiment. We were told to study and review patent of any DSP application an dthen implement the same. The signal processing was to be done on 1D signal. Signals like audio signal,ECG,EMG etc. are 1D signals. The application we selected was 'Audio Feature Extraction and Correlation'. The group members are Dhvanil Mandalia, Ameya Lokre, Chetan Jogi, Mohit Karandikar

The patent i reviewed was 'Audio Feature Extraction'

Patent No. : US20050232411 A1.

Summary:

Audio signature extraction, basically describes a method to detect a specific program or channel on basis of a unique key signature. This is used in television sets to identify various channel. The signature can be sent with data or when frame transmission is offline. The easiest method to implement this is using correlation. An incoming time domain moment can be captured and correlated with a stored reference signal to identify whether it belongs to a specific channel or not.

We implemented a system which stores time domian reference signal and correlates it with a test signal. If the correlation output is 1 then the signal is identified.

Link for the IEEE report, Plagiarism Report & US PATENT : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 9 - Basic Operations using DSP processor

This experiment was conducted in our classroom and we were taught the basic arithmethic and logical operations using the DSP processor by Mr. Jayganesh Rajaraman . He taught us the procedure of how to program a DSP processor , and also showed us various arithmethic and logical syntaxes . The lerning experience was amazing . We learnt addition, subtraction , multiplication , division and logical shift left,right etc. 

Experiment 8 - Digital FIR filter design using frequency sampling method (FSM)

From this experiment I learnt that phase response will be the same for LPF and HPF depending on whether the orders are kept same . Also the values of Ap and As have been verified .

link of the code : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 7 - Digital FIR filter design using windowing method

In this experiment we had to design a digital filter using windowing technique and study the spectrum of the filter .

From the experiment I learnt that in the magnitude response of the FIR  filter there we ripples present in the stopband of LPF . 

For BPF and LPF  the phase response is linear and the values of Ap and As almost match 

link of the code :https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 6 - Digital Chebyshev Filter Design

In this experiment we had to design a Digital Chebyshev Filter from Analog Chebyshev Filter BLT.

From the experiment I learnt that For a LP Digital Chebyshev Filter the ripples were present in the passband whereas the stopband was monotonic and vice versa for HP Digital Chebyshev Filter. Also in both the high pass and low pass filter, poles were present inside the unit circle , hence they are stable .For LPF there is definite zero at z=-1 while for HPF there is a definite zero at z=1.

The values of Ap and As as i/ps were approx. same . 

link of the code : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 5 - Digital Butterworth Filter design

In this experiment we had to design a digital filter from analog filter and study the aliasing effect due to sampling in Impulse Invariant Method and the frequency warping effect in BLT method .

So after obtaining the results i learnt that since observed values and i/p values are not matching , order of the filter had to be increased . The magnitude spectrum is monotonic in stopband and also monotonic in the passband for Digital Butterworth Filter. Analog LPF  poles lie on the LHS  of the s- plane and digial LPF  poles lie inside the unit circle . Hence both analog and digital butterworth LPF  are stable . 

link for the code : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 4 - Overlap Add Method (OAM) / Overlap Save Method(OSM)

In this experiment we had to implement filtering of long i/p sequence using OAM/OSM  algorithm .

After performing the experiment and observing the results I learnt that OAM and OSM are used for processing long data sequences .

Link for the code : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM 

Experiment 3 - Fast Fourier Transform

In this experiment we had to develop a program to perform FFT and IFFT of an N point signal . Here I considered N=4 

So for N=4 with x(n)={10,11,12,13 } , I obtained X[k]={46,-2+2j,-2,-2-2j} on performing FFT and vice versa for IFFT . 

So thus I learnt that the number of arithmetic calculations in FFT are very less as compared DFT . Hence FFT is a faster method . 

link for the code : https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 2 - Discrete Fourier Transform

In this experiment we had to perform DFT of 4 point and 8 point signal and also we had to conclude the effect of zero padding on magnitude spectrum . 

So after performing 4 point and 8 point DFT , what I observed was the as N increases , the frequency spacing reduces . Also the approximation error in representation of the spectrum decreases . Resolution of spectrum increases and hence visual appearance of spectrum quality improves . The missing values in 4 point DFT are present in 8 point DFT

Link of the code ;https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM

Experiment 1 - Convolution and Correlation Algorithms

The concept of Linear Convolution , Circular Convolution  and Correlation was taught in Signal and Systems . So from this experiment what we learnt that lenght of linear convoltuion o/p signal is N=L+M-1 where , L is the length of first i/p signal & M  is the length of the second i/p signal & N is the linear convolution o/p signal length .

As both the i/p signals are causal, the o/p is also causal.

Length of circular convolution o/p signal N=max(L,M) where L is the length of the first i/p signal and M is the length of the second i/p signal and N is the length of the o/p signal which is also the maximum of L and M .

For linear convolution using Circular convolution :- N>= L+M-1

Link of the Codes :https://drive.google.com/drive/folders/0BwYB93kRygZnT1E1MUNaZXh4RGM