[srslte-users] SRS support

Fri May 24 14:20:32 UTC 2019

Yes.. I did see the difference.. but I made changes appropriately.  Did you enable ue specific srs or cell specific??

Get Outlook for Android<https://aka.ms/ghei36>

________________________________
From: ALEJANDRO BLANCO PIZARRO <100283180 at alumnos.uc3m.es>
Sent: Friday, May 24, 2019 7:18:32 AM
To: Naveen Rawat; srslte-users
Subject: Re: [srslte-users] SRS support

Hi Naveen,

I do not remember such error when I did it. BTW, the version that I used of srsLTE was 17.10 I think.

Best,
Alejandro

El vie., 24 may. 2019 16:11, Naveen Rawat <naveen.rawat at cohere-technologies.com<mailto:naveen.rawat at cohere-technologies.com>> escribió:
Hi Alejandro,

After enabling SIB, I see that ping is not going through over LTE link between UE and eNB. Did you face this issue as well?

Thanks and Regards
Naveen Rawat

From: ALEJANDRO BLANCO PIZARRO <100283180 at alumnos.uc3m.es<mailto:100283180 at alumnos.uc3m.es>>
Date: Thursday, May 23, 2019 at 2:51 AM
To: Naveen Rawat <naveen.rawat at cohere-technologies.com<mailto:naveen.rawat at cohere-technologies.com>>
Subject: Re: [srslte-users] SRS support

Hi Naveen,

Thanks for response and this is good to know. Can you tell me to what extent were you able to make it work. I am worried that after SRS is sent by UE, since eNB cannot understand it, this could make the connection un-stable.

I do not completely understand it. Basically, the srsenb cannot handle the SRS, but this does not mean that the communication becomes unstable. What is going to happen, depending in your SRS configuration, you will have in a specific subframe and position the SRS and the srsenb is not going to anything with that information.  In my case, I saved the SRS in a file to later process it offline. I was interesting in estimating the distance between the UE and eNB but not in real time, so it is does not mind.

Please help me with following questions:

  1.  Location of your modifications.
What I am going to show you was personal decisions of the code, probably there is a better way to do it.

1º Add to the sib.conf (attached) the configuration regarding the SRS configuration:

srs_ul_cnfg =
{
    bw_cnfg = 0; // LIBLTE_RRC_SRS_BW_CONFIG_ENUM
    subfr_cnfg = 1; // LIBLTE_RRC_SRS_SUBFR_CONFIG_ENUM
    ack_nack_simul_tx = false; // bool
    max_up_pts = true; //  bool
    max_up_pts_present = true; // bool
    present = true; // bool
};

2º Add the lines to take the information from sib.conf in the parser: enb_cfg_parser (attached)

  // SRS_UL_CNFG configuration
  parser::section srs_ul_cnfg("srs_ul_cnfg");
  rr_config.add_subsection(&srs_ul_cnfg);

  srs_ul_cnfg.add_field(
    new parser::field_enum_num<LIBLTE_RRC_SRS_BW_CONFIG_ENUM,uint8>
    ("bw_cnfg", &data->rr_config_common_sib.srs_ul_cnfg.bw_cnfg,
    liblte_rrc_srs_bw_config_num,LIBLTE_RRC_SRS_BW_CONFIG_N_ITEMS)
  );
  srs_ul_cnfg.add_field(
    new parser::field_enum_num<LIBLTE_RRC_SRS_SUBFR_CONFIG_ENUM,uint8>
    ("subfr_cnfg", &data->rr_config_common_sib.srs_ul_cnfg.subfr_cnfg,
    liblte_rrc_srs_subfr_config_num,LIBLTE_RRC_SRS_SUBFR_CONFIG_N_ITEMS)
  );

  srs_ul_cnfg.add_field(
    new parser::field<bool>
    ("ack_nack_simul_tx", &data->rr_config_common_sib.srs_ul_cnfg.ack_nack_simul_tx)
  );
  srs_ul_cnfg.add_field(
    new parser::field<bool>
    ("max_up_pts", &data->rr_config_common_sib.srs_ul_cnfg.max_up_pts)
  );
  srs_ul_cnfg.add_field(
    new parser::field<bool>
    ("max_up_pts_present", &data->rr_config_common_sib.srs_ul_cnfg.max_up_pts_present)
  );
  srs_ul_cnfg.add_field(
    new parser::field<bool>
    ("present", &data->rr_config_common_sib.srs_ul_cnfg.present)
  );

And change from false to true this line: sib2->rr_config_common_sib.srs_ul_cnfg.present = true;

3º Add to srsenb/src/upper/rrc.cc these lines of code:

  // Configure Sounging Reference Signal
  if (parent->sib2.rr_config_common_sib.srs_ul_cnfg.present){
    fprintf(stderr, "Estamos configurnado srs\n");
    phy_cfg->srs_ul_cnfg_ded_present = 1;
    phy_cfg->srs_ul_cnfg_ded.srs_bandwidth = LIBLTE_RRC_SRS_BANDWIDTH_BW0;
    phy_cfg->srs_ul_cnfg_ded.srs_hopping_bandwidth = LIBLTE_RRC_SRS_HOPPING_BANDWIDTH_HBW0 ;
    phy_cfg->srs_ul_cnfg_ded.cyclic_shift = LIBLTE_RRC_CYCLIC_SHIFT_CS0 ;
    phy_cfg->srs_ul_cnfg_ded.freq_domain_pos = 0;
    phy_cfg->srs_ul_cnfg_ded.srs_cnfg_idx = 0;
    phy_cfg->srs_ul_cnfg_ded.tx_comb = 0;
    phy_cfg->srs_ul_cnfg_ded.setup_present = 1;
    phy_cfg->srs_ul_cnfg_ded.duration = 1;
  }

Now, the configuration from SIB2 and rrc messages supports SRS. As the UE decodes this informatio, it will send the SRS to the eNB without any configuration from the UE side.

4º Now you have to take the SRS from the eNB. What I did was to configure the SRS to be sent every 10 TTI, so I save the SRS to a file. Assuming that it is sent the last symbol of the subframe you can extract it. To do it, you can take it from the file phch_worker.cc, the variable where you can take it is the buffer of the eNB.
Here an example to do it:

srslte_filesink_t sink;
      int start = X % where the signal start
      int size = 576*2; % the size of the SRS
      srslte_filesink_init(&sink, buf, SRSLTE_COMPLEX_FLOAT_BIN);
      cf_t * signal = enb_ul.fft.out_buffer;
      srslte_filesink_write(&sink, (void *) &signal[start] , tot);

Sorry but this part is not programatical, it depends on your configuration. But was enough for my purpose.

  1.  With srsUE, you said SRS could be enabled, but srsEnb could not understand them. Did connection drop or did not even work?

     *   If srsENB somehow worked, did you make changes to ignore SRS received ?
The connection does not drop. And not

     *

  1.  Did you make progress with smartphone for SRS ?

Unfortunately not. Sorry

If you achieve to work with a smartphone tell me.

I totally recommend you to read the standard and forums to understadn how to configure the SRS.

Good luck. Best,
Alejandro

I am trying to achieve something similar to what you did, by using SRS at eNB for my experimentations.

Thanks and Regards
Naveen Rawat

From: ALEJANDRO BLANCO PIZARRO <100283180 at alumnos.uc3m.es<mailto:100283180 at alumnos.uc3m.es>>
Date: Wednesday, May 22, 2019 at 1:56 AM
To: Naveen Rawat <naveen.rawat at cohere-technologies.com<mailto:naveen.rawat at cohere-technologies.com>>
Cc: "srslte-users at lists.softwareradiosystems.com<mailto:srslte-users at lists.softwareradiosystems.com>" <srslte-users at lists.softwareradiosystems.com<mailto:srslte-users at lists.softwareradiosystems.com>>
Subject: Re: [srslte-users] SRS support

Hi Naveen,

>From my experience, SRS is not fully handled by srsLTE. As far as I know, srsUE is able to send it but srsenb cannot use it.

One year ago, I worked with SRS and basically what I did was to add some parameter to SIB2 and RRC messages to active them, thus the srsUE can send it. As the srsenb cannot handle it, I saved the SRS in a file and process it offline.

In addition, I tried to configure the SRS and check if a smartphone can send it, but I couldn't do it. I used the srsUE for my application.

Here you have the message that I posted in the mailing list of srsLTE: http://www.softwareradiosystems.com/pipermail/srslte-users/2018-May/001575.html. Thanks to Ismael I could use the SRS.

If you need further information tell me. I can tell you what was my configuration and which files of the code I modified.

Best,
Alejandro

El mar., 21 may. 2019 a las 9:10, Naveen Rawat (<naveen.rawat at cohere-technologies.com<mailto:naveen.rawat at cohere-technologies.com>>) escribió:
Hi srsLte Team,

I am looking for setting for enabling SRS. I found a bug was opened for this but looks like bug was closed with a commit. However I am unable to find the related config. Can you please tell if this was implemented from eNB side and what is the config to trigger that.

Thanks and Regards
Naveen Rawat
_______________________________________________
srslte-users mailing list
srslte-users at lists.softwareradiosystems.com<mailto: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/20190524/736a421c/attachment.htm>