[time-nuts] Effects of filter delay

Poul-Henning Kamp phk at phk.freebsd.dk
Sat Dec 23 18:34:57 EST 2006


In message <v04210100c1b362e32a0c@[10.0.0.3]>, Peter Vince writes:

>Please forgive what is probably a naive question from this newbie, 
>but a loop filter has considerable delay (half the number of samples 
>for a typical(?) symmetrical FIR or block-averaging filter).

One interesting approach here is to use a linear regression on the
last N measurements to predict where the next one will be.

This has slightly different results than a Kalman filter.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.



More information about the time-nuts mailing list