|
Authored by: Anonymous on Thursday, March 14 2013 @ 05:29 PM EDT |
...and since each algorithm is directed to a very specific method, and people
are able to calculate FFT other ways, each one does not preempt FFTs generally,
each one is patentable.
In Gottschalk v. Benson, the court declared that the claimed invention would
preempt all binary to bcd conversion. I'm pretty sure that was nonsense and I
know the decision is nonsense.
Hang your hat on Gottschalk v. Benson if you like, but the FFT patents just keep
coming:
Here's one that issued in January:
1. A method for area and speed efficient fast Fourier transform (FFT) processing
comprising:
mapping a one-dimensional discrete Fourier transform (DFT) to a
multi-dimensional representation;
re-indexing the multi-dimensional representation as a radix 23 decimation
architecture;
simplifying the radix 23 decimation architecture to obtain a nested butterfly
architecture;
acquiring N samples of a finite duration time-sampled signal; and
inputting the acquired N samples into the nested butterfly architecture to
obtain a N-point fast Fourier transform (FFT) output.
http://www.google.com/patents/US8346836?dq=fast+fourier+transform+method+2013&am
p;hl=en&sa=X&ei=tUBCUaLlGYHlyAGTxoGQAQ&ved=0CDMQ6AEwAA
How old is Gottschalk v. Benson?[ Reply to This | Parent | # ]
|
|
|
|
|