1 EXPERIMENT ONE: DOUBLE SIDEBAND MODULATION

1.1 INTRODUCTION
Double Sideband modulation is the easiest and most direct type of analog modulation. In this scheme, the
modulated signal is obtained using a direct multiplication of the modulating signal (i.e. the message) by a
cosine carrier. This multiplication results in shifting the entire spectrum of the message to a center
frequency defined by the carrier frequency. The modulation is said to be double sideband transmitted
carrier (DSB-TC) when the carrier is transmitted along the modulation term. If the carrier term is omitted,
the modulation is termed double sideband suppressed carrier (DSB-SC). DSB-TC has a significant
advantage in the receiver design (i.e. the envelop detector). Also transmitting the carrier independently
enables us to extract useful information such as the carrier frequency which can be helpful for carrier
synchronization. However, the DSB-TC loses to the other variant (i.e. the SC) in terms of power efficiency.

1.2 AIM
ed to achieve the following:

1. Get familiar with the concept of DSB modulation, and its parameters.
2. Study the performance of the DSB modulation.
3. Examine different detectors (coherent detector, envelope detector).
4. Study the performance of coherent detection in the presence of frequency or phase mismatch.

1.3 PROCEDURE
1. Use your PC to record a short (around 10-20 seconds in duration) audio signal. Use sampling
frequency = 48 KHz. Find the spectrum of this signal. It s important to record a short file or else
your PC will take FOREVER to run the simulation.
2. Using an ideal Filter, remove all frequencies greater than 4 KHz.
3. Obtain the filtered signal in time domain, this is a band limited signal of BW=4 KHz.
4. Find the mean square error (MSE) in the band limited signal. The MSE is defined as

Where, is the filtered voice signal, and x(t) is the authentic signal.

Note that this error should be small for voice signals. You could play the sound back, to make
sure only small distortion was introduced.

5. Modulate this carrier with the filtered signal you obtained, you are required to generate both
types of modulation. Choose a carrier frequency of 1MHz. For the DSB-TC, take the DC bias to be
twice the maximum of the message.

Note: The sampling frequency must be at least 2 times the carrier frequency, typical value
is . In this simulation, use .

dify the sampling frequency of the audio signal.

 6. For both types of modulations, use envelop detector to receive the message (assume no noise).
Note: to obtain the envelope you can use the following matlab command.

7. For both modulation types, find the error between the received message and the transmitted
message. Play the received signal back. What observation can you make of this?
8. For DSB-TC only, repeat steps 6-8 with SNR = 0, 10, and 20 dB. Play back the sound file each time
after detection. What conclusions do you make of that?
Note: To model the noise corruption step you can use the following identity

You can also use the matlab function awgn.

For DSB-SC, perform steps 9-11.
9. Use coherent detection to receive the signal and find the error, SNR=10 dB.
10. Repeat the coherent detection with frequency error, F=1.001MHz instead of 1MHz and Find the
error SNR=10dB. Do you have a name for this phenomenon?
11. Repeat the coherent detection with phase error = 100 and find the error, SNR=10db
12. Calculate the power efficiency for both types of modulation. Comment

1.4 USEFUL MATLAB FUNCTIONS

audiorecorder, fft, fftshift, ifft, awgn, upsample, downsample, resample, sound, Hilbert, abs, max.

1.5 REQUIREMENTS
Well commented matlab file.
Your obtained results (spectrum plots, error values, etc..) as well as your personal conclusions
should be included in a report.
2 EXPERIMENT TWO: SINGLE SIDEBAND MODULATION

2.1 INTRODUCTION
The bandwidth inefficiency stemming from the DSB transmission was the main reason why the single
sideband (SSB) was developed. In SSB modulation, the bandwidth required for band pass transmission is
equal to the bandwidth of that of the baseband. In other words, the band requirement for SSB is halved
with respect to that of DSB which requires twice the baseband bandwidth. This reduction in transmission
bandwidth is possible since the sideband are replicated twice over the positive and negative frequencies.
However, the bandwidth reduction doesn t come entirely free. In fact, SSB suffers from several
disadvantages that we hope to cover in the following simulation.

2.2 AIM
1. Familiarize students with the SSB modulation scheme.
2. Study the performance of SSB with different detectors.
3. Investigate the disadvantages of the SSB.

2.3 PROCEDURE
1. Use your PC to record a short (around 10-20 seconds in duration) audio signal. Use sampling
frequency = 48 KHz. Find the spectrum of this signal. It s important to record a short file or else
your PC will take FOREVER to run the simulation.
2. Using an ideal Filter, remove all frequencies greater than 4 KHz.
3. Obtain the filtered signal in time domain, this is a band limited signal of BW=4 KHz.
4. Find the mean square error (MSE) in the band limited signal. The MSE is defined as

Where, is the filtered voice signal, and x(t) is the authentic signal.

Note that this error should be small for voice signals. You could play the sound back, to make
sure only small distortion was introduced.

5. Generate a DSB modulated signal and plot its spectrum. Choose the carrier frequency to be
1MHz. Remember to set the sampling frequency to three times the carrier frequency.
6. Obtain the SSB by filtering out the USB of the DSB modulated signal and using an ideal
bandpass filter.
7. Using coherent detection with no noise interference, find the error in the received signal also
play the file back.
8. Repeat steps 6 and 7, only this time. Use a practical 3rd order Butterworth filter.
9. For the practical filter case, find the error in the received signal when SNR = 0, 10, and 20. Refer
to page 2 on how to model the additive white Gaussian noise. Play the sound back and leave
your comments.
 10. For the practical filter case and SNR = 10dB, assume coherent detection with
a. Frequency mismatch ( Take f = 1.001 MHz).
b. Phase mismatch (phase error = ).

What conclusion can you make based on your obtained results?

11. For the ideal filter case, generate a SSB-TC. As with experiment one, set the DC bias to twice
the maximum of the message. Use envelope detector to demodulate the message. Play back
on the received file and comment on the results obtained.
12. Can you list some of the disadvantages of SSB?
13. [Bonus] Can you generate the SSB signal in a way other than using DSB as an intermediate
step? Show this scheme and plot the resulting spectrum. [Hint: it involves the hilbert
command].

2.4 USEFUL COMMANDS

audiorecorder, filter, butter , fft, fftshift, ifft, awgn, upsample, downsample, resample, sound,
Hilbert, abs.

2.5 REQUIREMENTS
Well commented matlab file.
Your obtained results (spectrum plots, error values, etc..) as well as your personal conclusions
should be included in a report.
3 FREQUENCY MODULATION

3.1 INTRODUCTION
Frequency modulation (FM) is a modulation type in which the instantaneous frequency of the carrier is
changed according to the message amplitude. The motive behind the frequency modulation was to
develop a scheme with inherent ability to combat noise. The noise, being usually modeled as additive, has
a negative effect on the amplitude by introducing unavoidable random variations which are superimposed
on the desired signal. Unlike the amplitude, frequency has a latent immunity against noise. Since it resides
away from the amplitude, any changes in the amplitude would be completely irrelevant to the
frequency. In other words, there is no direct correlation between the variation in amplitude and
frequency, thus making FM a better candidate over AM with respect to noise immunity. However, what
FM gains in noise immunity lacks in bandwidth efficiency. Since FM usually occupies larger bandwidth,
AM is considered more bandwidth wise.

3.2 AIM
In this experiment, we investigate the narrowband frequency modulation when the SNR is varied.
tudents expected to:

1. Develop an appreciation of FM ability to counteract noise.

2. Be able to simulate the generation and the demodulation of NBFM using matlab.
3. To be able to tell the similarities and differences between AM and NBFM.

3.3 PROCEDURE
1. Repeat steps 1 through four in experiment 1.
2. Generate the NBFM signal. Use a carrier frequency of 1 MHz and a sampling frequency of
. Plot the resulting spectrum. What can you make out of the resulting plot?
3. Demodulate the NBFM signal using a differentiator and an ED. For the differentiator, you can
use the following command: diff. Assume no noise is introduced.
4. Find the mean square error in the received signal and play the sound back.
5. Repeat 3-4 for SNR = 0, 10, and 20 dB and leave your comments.

3.4 USEFUL COMMANDS

audiorecorder, filter, butter , fft, fftshift, ifft, awgn, upsample, downsample, resample, sound,
Hilbert, abs.

3.5 REQUIREMENTS
Well commented matlab file.
Your obtained results (spectrum plots, error values, etc..) as well as your personal conclusions
should be included in a report.