Re: [Openbts-discuss] B200 timing compensation

[Tom T. <tom...@et...>]

On Mon, Feb 8, 2016 at 2:57 AM, Ad Witblits <adw...@gm...> wrote:
> According to a commit made by Tom Tsou to Transceiver52M/UHDDevice.cpp on 15
> Apr 2015, some changes were made to the UHD 3.9.0 release that affected the
> timing values.

Correct. The UHD 3.9.0 included FIR filter changes to the FPGA
firmware; group delay of digital filters was correspondingly changed.
The referenced commit makes a build-time determination of UHD version
so that the appropriate timing values are used.

> Does anyone know what the mentioned loop-back test entail and how I can
> replicate the test for my B210 and B200?

There is currently no built-in mechanism to measure the
downlink-uplink timing offset, but the procedure is summarized below.

Timing is referenced from the downlink BCCH. For timing measurement in
loopback, the BCCH can be measured by tuning the Tx/Rx to the same
frequency and capturing the BCCH delay. Alternatively the RACH burst
of a close-range handset can be used.

I can provide rough patches to output timing measurements if there is interest.

> I want to update the uhd_dev_offset for OpenBTS-UMTS, where it hasn't been
> implemented.

UMTS is more complicated as the downlink-uplink implementations are
not symmetric. In this case, RACH burst measurement is more
appropriate.

  -TT