[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Patch-gnuradio] gr_rotator fix/optimization
From: |
Dan Halperin |
Subject: |
Re: [Patch-gnuradio] gr_rotator fix/optimization |
Date: |
Fri, 07 Sep 2007 14:41:03 -0700 |
User-agent: |
Thunderbird 1.5.0.13 (X11/20070824) |
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I ran some quick trials and found that C takes around 18k (I was
surprised it was this fast, actually!) iterations for the phaser
amplitude to be off by 0.1%. So K around a hundred or a thousand should
be good.
- -Dan
Matt Ettus wrote:
> Dan Halperin wrote:
>> Two changes:
>>
>> 1) Default phase increment changed from 1 to 0. This will not cause
>> faults and instead do nothing if the rotator phase is not initialized.
>
> Good idea.
>
>> 2) It seems like we can skip (expensive!) normalization of phase vector
>> if we simply normalize phase and increment upon setup. I suppose there
>> could be some lowest-order-bit precision errors that cause phase to
>> degrade over a super-long-running system, but we could surely at least
>> replace the current every-time normalization with normalization every K
>> calls, for large K (1 million+).
>
> I would keep the normalization -- it's in there because we ran into
> cases where the phasor really did decay to zero. However, I like the
> idea of doing it every K times. Even if K is 100 the performance impact
> will be negligible.
>
> Matt
>
> Matt
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFG4cVvy9GYuuMoUJ4RAtU7AJ9rmIPJ9/hW8Zo4V8vL9tbDWaztYQCcCFP6
+qCIWCEtkzja7XGBPYG1kzU=
=1EsY
-----END PGP SIGNATURE-----