The software-defined radio should have the following user-selectable
variables that can readily be set at the start ofprocessing of the received block of data:
- rolloff factor
for the square-root raised cosine pulse shape,
- initial phase offset,
- initial timing offset, and
- initial equalizer parameterization.
The following are some suggestions:
- Build your own transmitter in addition to
a digital receiver simulation.This will enable you to test your receiver
as described in the methodologyproposed in the preceding section
over a wider range of conditions than just thecases available on the website.
Also, building a transmitterwill increase your understanding of
the composition of the received signal.
- Try to break your receiver.
See how much noise can be present in the received signal beforeaccurate (e.g., less than 1% symbol errors) demodulation seems impossible.
Find the fastest change in the carrierphase that your receiver can track, even with a bad initial guess.
- In order to facilitate more effective debugging while building the project,
implementation of a debug mode in the receiver is recommended.The information of interest will be plots of the time histories
of pertinent signals as well as timinginformation (e.g., a graph of matched filter average output power
versus receiver symbol timing offset).One convenient way to add this feature
to your M
atlab receiver would be to include
a debug flag as an argumentthat produces these plots when the flag is activated.
- When debugging adaptive components, use a test
with initialization at the right answer and zero stepsize to determine if the problem is in the adaptation or in the fixed component structure.An initialization very near the desired answer with a small
stepsize will reveal that the adaptive portion isworking properly if the adaptive parameter trajectory
remains in the close vicinity of the desired answer.A rapid divergence may indicate that the update has the wrong sign
or that the stepsize is way too large.An aimless wandering that drifts away from the vicinity
of the desired answer represents a more subtle problemthat requires reconsideration of the algorithm code
and/or its suitability for the circumstance at hand.
Several test files that contain a “mystery signal”
with a quote from a well known bookare available on the website.
They are labelled
easyN.mat ,
mediumN.mat ,
and
hardN.mat .
These have been created with a variety of different
rolloff factors, carrier frequencies, phase noises,ISI, interferers, and symbol timing offsets.
We encourage the adventurous reader to try to“receive” these secret signals.
Solve the mystery.Break it down.
For further reading
An overview of a practical application of
software-defined radio emphasizing the
redefinability of the DSP portion of the receivercan be found in
- B. Bing and N. Jayant, “A cellphone for all standards,”
IEEE Spectrum , pg 34–39, May 2002.
The field of “software radio” erupted with a special issue of the
IEEE Communications Magazine in May 1995.
This was called a “landmark special issue”in an editorial in the more recent
- J. Mitola, III, V. Bose, B. M. Leiner, T. Turletti
and D. Tennenhouse, Ed.,
IEEE Journal on Selected
Areas in Communications (Special Issue on Software Radios),
vol. 17, April 1999.
For more information on the technological context and
the relevance of software implementations of communicationssystems, see
- E. Buracchini, “The Software Radio Concept,”
IEEE Communications Magazine , vol. 38,
pp. 138–143, September 2000
and papers from the (occasional) special section
in the
IEEE Communications Magazine on
topics in software and DSP in radio.For much more, see
- J. H. Reed,
Software Radio: A Modern Approach to
Radio Engineering , Prentice Hall, 2002
which overlaps in content (if not style) with the
first half of
Software Receiver Design .
Two recommended monographs that include more
attention than most to the methodologyof the same slice of digital receiver design as we
consider here are
- J. A. C. Bingham,
The Theory and Practice of Modem Design ,
Wiley Interscience, 1988.(especially Chapter 5)
- H. Meyr, M. Moeneclaey, and S. A. Fechtel,
Digital Communication Receivers: Synchronization, Channel
Estimation, and Signal Processing ,
Wiley Interscience, 1998(especially Section 4.1).
A similar design challenge can be found in the extensionto QAM found in the website in
A Digital Quadrature
Amplitude Modulation (QAM) Radio .
