[srslte-users] TTI handling PHY worker
Merlin Chlosta
merlin.chlosta+srslte-users at rub.de
Fri Jun 15 15:53:02 UTC 2018
Hi Jorge,
since you didn't get a reply the last time, I'll try to explain. Just be aware that I'm not familiar with these parts of the code.
When the continuous radio signal is received from the radio, srsLTE synchronizes to the cell using the LTE synchronization signals (PSS, SSS). After that, you no longer have some continuous data stream, but you exactly know the Physical Resource Block (PRB) boundaries.
I think [1] this is the code where the magic happens. If there's a new subframe available from radio (srslte_ue_sync_zerocopy_multi() == 1), give it to any available worker (i.e., call work_impl() on that SF). Well, and of course the radio only receives one subframe every millisecond. This implicitly creates the 1 ms step.
Merlin
[1] https://github.com/srsLTE/srsLTE/blob/77894ff7ef56238f57b2a7fb174c13fc91c3dc5a/srsue/src/phy/phch_recv.cc#L634
On 15.06.2018 00:26, Jorge Gomez Velasquez wrote:
> Dear all,
>
>
> From my understanding at **phch_worker::work_impl()** the UL and DL scheduling takes place (MAC functions are called there). I would expect the function *phch_worker::work_impl()* to be called each millisecond.
>
> I see that this function (thread) is indirectly called from the*txrx::run_thread()* function. Where and how is this one millisecond time step managed ?
>
>
> Thanks in advance
>
>
> Jorge Gomez
>
>
>
>
> _______________________________________________
> srslte-users mailing list
> srslte-users at lists.softwareradiosystems.com
> http://www.softwareradiosystems.com/mailman/listinfo/srslte-users
>
More information about the srsran-users
mailing list