Dear Mande
here is an application that requires significant FPGA
functionality:
try to design the first stages of a spread-spectrum receiver,
ie, the part of the receiver that does pn code acquisition
(and tracking). This is a demanding task that needs to be performed on
hardware. Once code acquisition/tracking is performed, the input data
can be despread (this is also done in hardware) and then only the
symbol-spaced samples are passed to the Gnu-radio software for
further processing (ie, decoding, etc).
Some receivers require 2 or 3 such blocks (fingers) running in parallel.
Check out how IS-95 mobile standard works for more info.
Achilleas
2007/6/15, Achilleas A. [email protected]:
symbol-spaced samples are passed to the Gnu-radio software for
further processing (ie, decoding, etc).
I do not know the rate of the spreading codes used in the IS-95
standard, but several people have made complete or parts of GPS
receivers for GNU Radio, where all of the processing is done in
real-time. The lack of hardware multipliers in the Cyclone I is the
biggest limiting factor for creating the these correlators in the
FPGA.
Some receivers require 2 or 3 such blocks (fingers) running in parallel.
Check out how IS-95 mobile standard works for more info.
In a GPS receiver, it is desirable to be able to run at least 4
channels in parallel. This highly specialized functionality is better
suited for software. With modern CPUs, SIMD instructions and advanced
algorithms, I do not think the performance here is that much
different.
–
Trond D.