Scribd Logo
Upload

Welcome to Scribd!

  • 5 views

    Bio Assingnment 1 Codes

    Uploaded by

    Thatukuru Lakshman
    BIOLOGY MEDICAL ASSIGNMENT

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Bio Assingnment 1 Codes

    Uploaded by

    Thatukuru Lakshman
    5 views13 pages
    BIOLOGY MEDICAL ASSIGNMENT

    Original Title

    Bio_assingnment_1_codes

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    BIOLOGY MEDICAL ASSIGNMENT

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    5 views13 pages

    Bio Assingnment 1 Codes

    Uploaded by

    Thatukuru Lakshman
    BIOLOGY MEDICAL ASSIGNMENT

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    You are on page 1of 13

    BIO-MEDICAL SIGNAL PROCESSING

    ASSIGNMENT – 1

    - Thatukuru Lakshman
    - CB.EN.U4AIE22067

    6) Aliasing Demonstration in Matlab

    clc;

    clear all;

    close all;

    samplingFrequency = 50;

    signalFrequency = 10;

    frequencyShiftFactor = 100;

    shiftedFrequency = signalFrequency + frequencyShiftFactor *


    samplingFrequency;

    timeVector = 0:1/samplingFrequency:1;

    signal1 = sin(2 * pi * signalFrequency * timeVector);

    signal2 = sin(2 * pi * shiftedFrequency * timeVector);

    figure;

    subplot(2, 1, 1);

    stem(timeVector, signal1, 'filled');

    title('Signal with original frequency');

    subplot(2, 1, 2);

    stem(timeVector, signal2, 'filled');

    title('Signal with shifted frequency');


    OUTPUT

    7) Demonstration of Power Spectrum and Power Spectral Density(PSD) in


    Matlab

    clc;

    clear all;

    close all;

    fs = 500;

    N = 1000;

    x = 0:N-1;

    t = x * (1/fs);
    y = 1 * sin(2 * pi * 100 * t) + 4 * sin(2 * pi * 110 * t);

    Y = fft(y);

    Power_spectrum = (abs(Y) / N).^2;

    f = (0:N-1) * (fs / N);

    Power_spectrum = Power_spectrum(1:N/2+1);

    f = f(1:N/2+1);

    PSD = Power_spectrum / N;

    figure;

    subplot(3,1,1);

    plot(t, y);

    title('Original Signal');

    xlabel('Time (s)');

    ylabel('Amplitude');

    grid on;

    subplot(3,1,2);

    plot(f, 10*log10(Power_spectrum));

    title("Power Spectrum of the Signal");

    xlabel('Frequency (Hz)');

    ylabel('Power/Frequency (dB/Hz)');

    grid on;

    subplot(3,1,3);

    plot(f, 10*log10(PSD));

    title("Power Spectral Density of the Signal");

    xlabel('Frequency (Hz)');

    ylabel('Power/Frequency (dB/Hz)');

    grid on;
    OUTPUT

    9) Demonstration of a signal after Interpolation in Maltab

    clc; clear all; close all;

    fs = 50; % Original sampling frequency (Hz)

    % Load the eeg data from a .mat file

    data = load("/Users/lakshmanthatukuru/Desktop/eeg_data.mat");

    y1 = data.eeg; % Extract the eeg data from the loaded structure

    N = length(y1); % Number of samples in the original signal

    t = (0:N-1)/fs; % Time vector for the original signal


    new_fs = 100; % New (higher) sampling frequency for interpolation

    t1 = t(1):1/new_fs:t(end); % Time vector for the interpolated signal

    % Performing cubic spline interpolation

    new_y = interp1(t, y1, t1, 'cubic');

    % Plotting

    figure;

    % Original signal

    subplot(2,1,1);

    plot(t, y1, 'r');

    title('Signal before interpolation');

    xlabel('Time (s)');

    ylabel('Amplitude');

    % Interpolated signal

    subplot(2,1,2);

    plot(t1, new_y, 'b');

    title('Signal after Interpolation');

    xlabel('Time (s)');

    ylabel('Amplitude (Interpolated)');

    OUTPUT
    10) The file heart_rate_sample contains 2 heart rates – hr_med and hr_pre
    during two conditions with corresponding time vectors t_med and t_pre.
    Both these signals are unevenly sampled. Convert the hr_pre to evenly
    spaced data (resample) using interpolation. You can take the sampling
    frequency as 100Hz

    clc; clear all; close all;

    % Load the heart rate data from a .mat file

    data = load ("/Users/lakshmanthatukuru/Desktop/heart_rate_samples.mat");

    hr_pre = data.hr_pre;

    t_pre = data.t_pre;

    t = t_pre

    N = length(hr_pre) % Number of samples in the original signal

    new_fs = 100; % New (higher) sampling frequency for interpolation

    t1 = t(1):1/new_fs:t(end) ; % Time vector for the interpolated signal

    % Performing cubic spline interpolation

    new_y = interp1(t_pre, hr_pre, t1, 'cubic');


    % Original signal

    figure;

    subplot(2,1,1);

    plot(t, hr_pre, 'r');

    title('Signal before interpolation');

    xlabel('Time (s)');

    ylabel('Amplitude');

    % Interpolated signal

    subplot(2,1,2);

    plot(t1, new_y, 'b');

    title('Signal after Interpolation');

    xlabel('Time (s)');

    ylabel('Amplitude (Interpolated)');

    OUTPUT
    11) Create a noisy signal to generate a waveform of 200,400 Hz in
    MATLAB with SNR -8 dB, sampling frequency=1000Hz, and N=512
    samples. Plot magnitude spectrum.

    Repeat with -4dB, -16dB, plot the magnitude spectrums – write your
    observation in detecting the signal frequency.
    Repeat with N =1000 and N=200 samples, plot the magnitude spectrums -
    and write the inference when detecting the signal frequency.

    clc; clear; close all;

    fs = 1000; % Sampling frequency in Hz

    N = 512; % Number of samples

    f1 = 200; % Frequency of the first sine wave in Hz

    f2 = 400; % Frequency of the second sine wave in Hz

    % Time vector

    t = (0:N-1) / fs;

    % Signal generation: sum of two sine waves

    y = sin(2 * pi * f1 * t) + sin(2 * pi * f2 * t);

    % SNR values to test

    snr_values = [-4, -8];

    % Plotting for different SNR values

    figure;

    for i = 1:length(snr_values)

    snr = snr_values(i);

    % Signal power

    signal_power = mean(y .^ 2);

    % Noise power calculation from SNR

    noise_power = signal_power / (10 ^ (snr / 10));


    % Generate Gaussian noise

    noise = randn(size(y)) * sqrt(noise_power); % Noise with given power

    % Add noise to the signal

    signal = y + noise;

    % FFT and normalization

    fft_sign = abs(fft(signal) / N); % Normalize the FFT result

    % Take the positive half of the spectrum

    fft_sign = fft_sign(1:N/2+1);

    f = (0:N/2) * (fs / N);

    % Plot magnitude spectrum

    subplot(2, 1, i);

    plot(f, fft_sign, 'c');

    xlabel('Frequency (Hz)');

    ylabel('Magnitude');

    title(['SNR = ', num2str(snr), ' dB, N = ', num2str(N)]);

    end

    % Repeat with different number of samples

    N_values = [1000, 200];

    for j = 1:length(N_values)

    N = N_values(j);

    t = (0:N-1) / fs;

    y = sin(2 * pi * f1 * t) + sin(2 * pi * f2 * t);

    figure;

    for i = 1:length(snr_values)

    snr = snr_values(i);

    % Signal power
    signal_power = mean(y .^ 2);

    % Noise power calculation from SNR

    noise_power = signal_power / (10 ^ (snr / 10));

    % Generate Gaussian noise

    noise = randn(size(y)) * sqrt(noise_power); % Noise with given


    power

    % Add noise to the signal

    signal = y + noise;

    % FFT and normalization

    fft_sign = abs(fft(signal) / N); % Normalize the FFT result

    % Take the positive half of the spectrum

    fft_sign = fft_sign(1:N/2+1);

    f = (0:N/2) * (fs / N);

    % Plot magnitude spectrum

    subplot(length(snr_values), 1, i);

    plot(f, fft_sign, 'c');

    xlabel('Frequency (Hz)');

    ylabel('Magnitude');

    title(['SNR = ', num2str(snr), ' dB, N = ', num2str(N)]);

    end

    end

    OUTPUT
    Figure 1:
    Figure 2:
    Figure 3:

    The interpretation you've provided explains how variations in Signal-to-Noise


    Ratio (SNR) and sample size (N) affect the clarity and resolution of signal peaks
    in the frequency domain:

    1. SNR Variation:
    - Lower SNR values: When the SNR is low, the noise level is high relative to
    the signal. This increased noise makes it more challenging to discern the true
    signal components, resulting in less clear or visible peaks at specific frequencies
    like 200 Hz and 400 Hz. The signal is overwhelmed by noise, which obscures
    the important features of the signal.
    - Higher SNR values: With a higher SNR, the signal stands out more clearly
    above the noise. This makes the signal peaks at 200 Hz and 400 Hz more
    distinguishable and easier to detect. The reduced noise interference allows for a
    clearer representation of the signal in the frequency domain.

    2. Sample Size Variation:


    - Increasing the number of samples (N): A larger sample size improves the
    frequency resolution of the signal. This means the frequency components of the
    signal can be more precisely identified, leading to sharper and more distinct
    peaks at 200 Hz and 400 Hz. The signal's frequency content is captured more
    accurately, resulting in better-defined peaks.
    - Fewer samples: With a smaller sample size, the frequency resolution
    decreases. This causes the signal peaks to appear broader and less defined, as
    there are fewer data points to accurately represent the frequency content. The
    reduced sample size limits the ability to precisely identify the frequencies
    present in the signal.

    This interpretation highlights the importance of both SNR and sample size in
    accurately analyzing signals, particularly in applications like signal processing
    and spectral analysis..

    You might also like