GNU SDR DSP backend question

I am currently working at Princeton University in a department
responcible
for some fMRI dsp work and managing a cluster for image processing.

We are currently discussing new projects to use the cluster for
including a
new tool for performing the different comparisons and algorithms in an
easy
to use implementation. It appears that what we are looking for is very
similar to the back end architecture GNU SDR uses (the blocks and swig
wrapping section, if not the actual transmission of data). If somebody
could link me some information on how the block code architecture was
concieved or (assuming you based it on a previously exsisting setup)
some
information as to where to start with a design of this type I’d be very
happy. This type of technology could be very useful in the field of
fMRI
and I hope to hear from you soon.
-Garrett McGrath

On Thu, 2 Nov 2006, Garrett McGrath wrote:

If somebody
could link me some information on how the block code architecture was
concieved …

The JSAC96 and JSAC98 papers linked below are informative.
As I understand it, the VuSystem and SpectrumWare/PSpectra
projects at MIT in the mid and late 1990’s influenced GNU
Radio. Early GNU Radio distributions included the PSpectra code, I
recall. According to the papers, VuSystem was the immediate ancestor
of
PSPectra.

VuSystem, multimedia
http://tns-www.lcs.mit.edu/vs/vusystem.html
http://tns-www.lcs.mit.edu/publications/JSAC96.html

SpectrumWare/PSpectra, software radio
http://www.sds.lcs.mit.edu/SpectrumWare/home.html
http://www.sds.lcs.mit.edu/~vanu/JSAC98.ps

Jon Jacky