ELEC 470 Class Notes and Demonstrations, Spring 1998


Monday, February 9: Hilbert transform hilbert_demo.m

Friday, February 13: Amplitude Modulation in Simulink
(See Homework 5 for questions related to these programs.)

Run the following Simulink programs in matlab5 on the Sun computers. To start filename.mdl, type filename at the Matlab prompt.

DSB-SC demodulation using Costas phase-locked loop:
  1. Download the programs costas_init.m and costas_demo.mdl
  2. Start matlab5 on the Sun, and type costas_init followed by costas_demo at the MATLAB prompt.


Wednesday, March 4, 1998:

MATLAB program mod_ex.m for analog modulation and demodulation for Project 2.
(Program was developed and tested using matlab5 on the Sun computers.)


Monday, March 30, 1998: Quantization Demonstration

MATLAB program quant_demo.m
Try changing the number of bits to R = 7, 6, 5, 4, 3, 2, 1. At which number of bits and corresponding signal-to-quantization-noise ratio does the noise due to quantization become audible?

If you just want the results, you can click below to hear the effects of quanitizing to R bits or L = 2^R + 1 levels:

Original, 7 bits, 6 bits, 5 bits, 4 bits, 3 bits, 2 bits, 1 bit.


Monday, April 6, 1998: Matched Filter Demonstration

MATLAB program matchfilt.m


Monday, April 13, 1998: Raised Cosine Zero-ISI Demonstration

MATLAB program rcdemo.mdl
  1. Run the Simulink program rcdemo.mdl in matlab5
  2. Try changing the "Filter rolloff factor" in the "R-C filter" block. This is the "alpha" parameter of the raised-cosine pulse, and the allowable range is from 0 to 1. Recall that alpha = 0 is a sinc-shaped pulse.
  3. Observe the eye diagram for different values of alpha. Is the eye "more open" when alpha = 1? Can you see much less sensitivity to errors in the sampling times when alpha is close to 1?
  4. Try increasing the noise size by changing "sigma" in the "Rayl N" block from 0.001 to 0.1. See how the "scatter diagram" smears, and also see the effect on the eye diagram.
  5. Be sure that you understand how the eye diagram and scatter diagram are produced!


Friday, April 17, 1998: The effect of bit errors on speech in a 64 kbps PCM system

Here are three speech files to listen to. They are designed to illustrate how bit errors in a PCM system affect the quality of speech that is heard at the receiver. Each file is produced by sampling speech 8000 times per second and quantizing each sample with 8 bits.
  1. clint.au is the original speech signal, with no transmission errors.
  2. c1.au corresponds to transmission over a channel with a bit error rate (BER) of 0.1, so that 1 in 10 bits is received incorrectly (on average).
  3. c4.au corresponds to transmission over a channel with a BER of 0.001, so that 1 bit in every 1000 bits is received incorrectly (on average).

Amati Communications Corporation and ADSL (Asymmetric Digital Subscriber Line)
Look at the "white paper" for a technical description of ADSL.


Monday, April 20, 1998: Digital Communication System Design Exercise


April 22, 24, and 27, 1998: Adaptive Equalization

  1. MATLAB program zfeq.m for Zero-Forcing (ZF) Equalization with no noise

    Try changing the channel time constant tau in the program. Notice how the equalizer "opens the eye" perfectly!

  2. MATLAB program zfeqn.m for Zero-Forcing Equalization with additive noise

    Notice that the equalizer seems to amplify the noise: the noise looks larger in the eye diagram after equalization. Why is this true?

  3. MATLAB program wienereq.m for Wiener Filter Equalization with additive noise. This program includes the ZF equalizer for comparison.

  4. MATLAB program lmseq.m for adaptive equalization using the LMS algorithm.


May 1, 1998: Passband Digital Modulation

Please see Project 3

Also see Links to CDMA sites.