[srslte-users] Low Leve PHY sync method usage

[Sylvain Munaut]

Hi Andre,


> $ ./lib/src/phy/sync/test/pss_usrp -f 806000000 -i 257 -s 64

I'll give that example a try and read it see if I can gather some info from it.


> with both an RTL SDR and a USRP and both work fine. Note that the PSS is
> only in the center 6 PRB, so a small FFT size is enough.

Sure but :

(1) I have samples coming in at 30.72Msps, don't have a choice there
(2) I need the alignement as precise as possible at that rate

If there is another way to get this in srsLTE, I'm all ear.
I see there are some variants of the functions with 'decimate'
parameters, maybe there is a way to feed it 30.72M at the input but
tell it to only look at the center to do a coarse search and then
refine it, maybe when doing the SSS ?

But I couldn't find any documentation whatsoever for those low level
functions describing what they do or how to properly use them which
makes it pretty hard to find out the best way to achieve something.
And their name and argument are really not always that explicit.

For instance,  srslte_pss_init_fft takes a frame_size argument, but
turns out it's actually half the frame size you need to give it
AFAICT.
Also, I'm still not sure exactly what the offset that's return
corresponds to. It seems to be to the first sample after where the PSS
is found, which isn't exactly obvious (I would have expected the first
sample of the PSS or the first sample of the frame/half-frame).


Cheers,

    Sylvain