[srslte-users] Can't receive data in subframe 0

Tue Apr 11 16:36:17 UTC 2017

Hi all,

I had a quick discussion with Felipe, and it seems that part of the problem
was a little miss-comprehension on how the Physical channels were used.

Until now I've always been able to allocate a DCI message on the PCCH data
on the PDLSCH (with 6 RBs and 3 CFI); at PHY layer, if I'm not wrong, the
PCCH act as an indexing channel. Basically it provides the necessary
information to read the actual data, but you should not write MAC/RRC
control data on the PCCH, since it is multiplexed in the PDLSCH. Can be
pictured as the "index for a page made of paragraphs".

At high values of the CFI, potentially one can have a higher granularity in
the allocation of single RB or RB groups, while with reduced CFI values(1),
you have less DCI slots and so less granularity on how to organize your
sub-frame data.

There is an additional question I have for srsLTE gurus or everyone else
here:
usually I select a Resource Allocation(RA) strategy (type 0, 1 or 2) to
create the DCI message, but once I pick one RA type, am I restricted to
keep using that in the sub-frame?
Technically, If I introduce a way to avoid RB allocation collision (a DCI
cannot reserve RB of another DCI), I don't see anything which restrict me
to do so... any comment about this?

Cheers,
Kewin R.

On Tue, Apr 11, 2017 at 6:11 PM, Ismael Gomez <
ismael.gomez at softwareradiosystems.com> wrote:

> Hi all, I think Felipe might be right and there might be some combination
> of CFI and DCI allocation that is undetected by the pdcch_ue_locations
> function. I rewrote some bits of that function to make it more clear and
> readable and added the suggestion from Felipe.
>
> would be great if you could try it. I just pushed it on next.
>
> Thanks for catching that
>
> On Tue, 11 Apr 2017 at 14:20 Kewin Rausch <krausch at fbk.eu> wrote:
>
>> Hi everyone,
>>
>> I'm currently able to transmit data on subframe 0 without having to
>> modify the srsLTE library code.
>> I've already tested such data transmission with srsUE and legacy hardware
>> like Nexus 5, and it works nicely (data transmitted on subframe 0, in my
>> case, is the SIB2).
>>
>> My CFI is 3, and I use 6 RB (1.4 MHz) LTE signal.
>>
>> Cheers,
>> Kewin R.
>>
>>
>> On Tue, Apr 11, 2017 at 1:57 PM, Felipe Augusto Pereira de Figueiredo <
>> zz4fap at gmail.com> wrote:
>>
>> Hi Folks,
>>
>> I ran some tests and figured out DCI was not being found during the DCI
>> search.
>> Then, I went to function "srslte_pdcch_ue_locations_ncce" in pdcch.c and
>> changed the following line. The "for" should follow the values in Table
>> 9.1.1-1: PDCCH candidates monitored by a UE in TS 36.213.
>>
>>     //for (i = 0; i < SRSLTE_MIN(nof_cce / L,
>> S[l]/PDCCH_FORMAT_NOF_CCE(l)); i++) {
>>     for (i = 0; i < (S[l]/PDCCH_FORMAT_NOF_CCE(l)); i++) {
>>
>> Now I'm able to find the CRC corresponding to the C-RNTI being used,
>> however, I still can't decode the transmitted data.
>>
>> Kind Regards,
>>
>> Felipe Augusto
>>
>> On Tue, Apr 11, 2017 at 10:18 AM, Ismael Gomez <ismael.gomez@
>> softwareradiosystems.com> wrote:
>>
>> sharetechnote.com is a good reference.
>>
>> On Mon, 10 Apr 2017 at 16:18 Felipe Augusto Pereira de Figueiredo <
>> zz4fap at gmail.com> wrote:
>>
>> Dear folks,
>>
>> Could you kindly provide me with some references on how to calculate if
>> there is enough space for the DCI and the PDSCH for a given CFI
>> configuration?
>> As told before, I'm trying to transmit data on subframe 0 but it has been
>> no avail so far.
>>
>> I'm using CFI equal to 1 instead of 3 (but I also tried with CFI=3),
>> sending just one DCI (I'm using the pdsch_enodeb example) and the data is
>> using all the RBs (25 RBs) with MCS=0 which corresponds to a TB of 680 bits.
>>
>> I'd like to calculate it and try to understand what is happening.
>>
>> Many thanks and Best regards,
>>
>> Felipe Augusto
>>
>>
>>
>> On Mon, Apr 3, 2017 at 10:50 PM, Ismael Gomez <ismael.gomez@
>> softwareradiosystems.com> wrote:
>>
>> I don't know. You should calculate it. It depends on the BW.
>>
>> On Mon, 3 Apr 2017 at 22:49 Felipe Augusto Pereira de Figueiredo <
>> zz4fap at gmail.com> wrote:
>>
>> Dear Ismael,
>>
>> Many thanks for your reply!
>>
>> I'm using CFI equal to 1 instead of 3 (but I also tried with CFI=3),
>> sending just one DCI (I'm using the pdsch_enodeb example) and the data is
>> using all the RBs (25 RBs) with MCS=0 which corresponds to a TB of 680 bits.
>>
>> Do you think I have enough space for the DCI with that configuration?
>>
>> Thanks and Kind Regards,
>>
>> Felipe Augusto
>>
>>
>>
>> On Mon, Apr 3, 2017 at 10:41 PM, Ismael Gomez <ismael.gomez@
>> softwareradiosystems.com> wrote:
>>
>> Hi Felipe,
>>
>> There are less resource elements in subframe 0 for PDSCH. Make sure that
>> there is enough space for the DCI and the PDSCH for your CFI configuration.
>>
>> On Sun, 2 Apr 2017 at 11:16 Felipe Augusto Pereira de Figueiredo <
>> zz4fap at gmail.com> wrote:
>>
>> Hi Folks,
>>
>> I'm doing some experiments with the srsLTE source code and I'd like to
>> transmit data on every subframe 0, however, even after modifying the source
>> code (the pdsch enodeb and ue examples) I'm not receiving anything on any
>> subframe 0 except PBCH transmissions (MIB). I added some printf's and
>> figured out that no DCI is being found during the blind search at ue side.
>>
>> I'm not sure if the problem is that the enodeb is not adding the DCI
>> properly or if the ue is not being able to decode DCI on subframe 0.
>>
>> Is there any bug/tricky that refrains the eNodeB or the UE from
>> transmitting/receiving data on subframe 0?
>>
>> I'm using pdsch_enodeb and pdsch_ue examples.
>>
>> Many thanks in advance and Kind Regards,
>>
>> Felipe Augusto
>>
>>
>>
>> ---------- Forwarded message ----------
>> From: *Felipe Augusto Pereira de Figueiredo* <zz4fap at gmail.com>
>> Date: Fri, Mar 31, 2017 at 5:17 PM
>> Subject: Can't receive data in subframe 0
>> To: srslte-users at lists.softwareradiosystems.com
>>
>>
>> Hi srsLTE group,
>>
>> I'm doing some experiments with the code and I'd like to transmit data on
>> subframe 0, however, even after modifying the source code I'm not receiving
>> anything on any subframe 0 except PBCH transmissions (MIB). I figured out
>> that any DCI is being found during the blind search but as I understand, at
>> eNodeB side it is being added to the grid.
>>
>> Is there any bug/tricky that refrains the eNodeB or the UE from
>> transmitting/receiving data on subframe 0?
>>
>> I'm using pdsch_enodeb and pdsch_ue examples.
>>
>> Thanks and Kind Regards,
>>
>> Felipe Augusto
>>
>>
>>
>>
>>
>> _______________________________________________
>> srslte-users mailing list
>> srslte-users at lists.softwareradiosystems.com
>> http://www.softwareradiosystems.com/mailman/listinfo/srslte-users
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20170411/6ce35dfc/attachment.htm>