Assignments

Demo examples

  1. A Sampled Spectrum

    Source files in spectrum_sampling.zip

    Just as a time-domain signal, a discrete spectrum can be over- or undersampled. DFT method inherently produces a sampled (i.e. discrete) spectrum, the continuous analog of which would be found with the use of discrete-time Fourier transform (DTFT).

    Examine the program that demonstrates spectrum undersampling and how the signal is interpolated to densen the spectrum.

  1. Real and complex mixing

    Source files in mixing.zip

    In this example, we will demonstrate the concept of frequency upconversion/downconversion with an NCO in digital domain. Example illustrates how input signal and end-result of downconversion/upconversion look like in time and frequency domain.

  1. Filtering

    Source files in filtering.zip

    We demonstrate basic types of FIR filters: low-pass, high-pass and band-pass. The filters and signals they operate on are presented in time and frequency domain.

  1. Multirate processing

    Source files in multirate.zip

    In this example, effects of signal decimation and interpolation are presented in time and frequency domain.

  1. 4-QAM modulator

    Source files in modulations.zip

    This example demonstrates 4-QAM modulation using a sine-wave carrier. The modulated signal is presented in time and frequency domain.

  1. OFDM symbol generator

    Source files in ofdm_tx.zip

    In this example, we will demonstrate how the OFDM symbol is generated on transmitter side. Example illustrates how input bitstream is coded with QAM16 and incorporated into OFDM subcarriers

  1. From Antennas to Arrays

    Source files in antennas.zip

    To explain the design of antenna arrays, we look at how we start off from a current distribution (a line source) to get the array factor. We then discretize it to get the beamforming coefficients for each antenna element.

Self-study assignments

Assignment 01: A Quantized Ode

Source files in quantization.zip

You are given an audio file which is oversampled and overquantized. This means it can be reduced to a lower sample rate and a smaller number of bits per sample, while retaining all the data and not reducing SNR.

Produce a program which optimizes the given audio file.

Note

Consider the cutoff frequency to be the one at which FFT spectrogram shows -60 dB intensity.

Input Files
  • odeToJoy_oversampled.wav - The input audio file.
Output Files
  • odeToJoy_optimized.wav - The optimized audio file.
  • Your solution as a MATLAB or Python script.

Assignment 02: OFDM symbol demodulator

Source files in ofdm_rx.zip

You are provided with I_samples.csv and Q_samples.csv files, in which I and Q time domain samples are stored in floating-point format. These I/Q samples are the result of OFDM symbol generation, done as presented in OFDM symbol generator demo example.

Your task is to perform OFDM demodulation and obtain 16-QAM symbols encoded in the OFDM symbol. Your solution must be stored in a rx_symbols.bit binary file, where each 16-QAM symbol is stored as a uint8 value.

Input Files
  • I_samples.csv - In-phase time-domain input samples.
  • Q_samples.csv - Quadrature-phase time-domain input samples.
Output Files
  • rx_symbols.bit - Recovered 16-QAM symbols.
  • Your solution as a MATLAB or Python script.