SRS periodicity VS Number of UE. How does it work?

['Marco Pennacchioni' via ns-3-users]

Hi everyone!
I'm Marco, I'm developing a master thesis on NB-IoT and LTE systems. I've a
question about the relation beetween the SRS periodicity and the number of
UEs in a LTE simulation. I need to simulate a scenario including 50000 UEs
and, reading in a previous thread the problems due to the value of SRS
periodicity, I estimated and implemented SRS values over 640 (note that
3GPP specification stops at SRS periodicity =320), in the following files:


*lte-phy.cc , lte-enb-rrc.cc :*


/**
* Sounding Reference Symbol (SRS) periodicity (TSRS) in milliseconds. Taken
* from 3GPP TS 36.213 Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsPeriodicity[SRS_ENTRIES] = {0, 2, 5, 10, 20, 40,
80, 160, 320, 640, 1280, 2560, 5120, 10240, 20480, 40960, 81920};
/**
* The lower bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiLow[SRS_ENTRIES] = {0, 0, 2, 7, 17, 37,
77, 157, 317, 637, 1277, 2557, 5117, 10237, 20477, 40957, 81917};
/**
* The upper bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiHigh[SRS_ENTRIES] = {0, 1, 6, 16, 36, 76,
156, 316, 636, 1276, 2556, 5116, 10236, 20476, 40956, 81916, 163836};


The question is: doing that the simulation works, but I found an anomaly.
Setting a value of SRS periodicity = 640, the system works even if we put a
number of UEs greater than this value. For example, I choose 1000 UEs and
SRS periodicity = 640. I would expect this not to work. Does anyone know
why this happens? Is this value of SRS too big, so that it becomes
meaningless?

Thanks in advance

Marco Pennacchioni

[Marco Miozzo]

Hi,

in principle you should have one SRS periodicity per each UE attached to
(in order to make them able to send SRS signal), so maybe not all the the
UEs are able to attach in your simulation, since in that case an ASSERT
should be triggered. You can check it with the MAC and/or RLC stats.

Best,
marco.


On Wed, Mar 29, 2017 at 3:15 PM, 'Marco Pennacchioni' via ns-3-users <

Post by 'Marco Pennacchioni' via ns-3-users
Hi everyone!
I'm Marco, I'm developing a master thesis on NB-IoT and LTE systems. I've
a question about the relation beetween the SRS periodicity and the number
of UEs in a LTE simulation. I need to simulate a scenario including 50000
UEs and, reading in a previous thread the problems due to the value of SRS
periodicity, I estimated and implemented SRS values over 640 (note that
*lte-phy.cc , lte-enb-rrc.cc :*
/**
* Sounding Reference Symbol (SRS) periodicity (TSRS) in milliseconds. Taken
* from 3GPP TS 36.213 Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsPeriodicity[SRS_ENTRIES] = {0, 2, 5, 10, 20,
40, 80, 160, 320, 640, 1280, 2560, 5120, 10240, 20480, 40960, 81920};
/**
* The lower bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiLow[SRS_ENTRIES] = {0, 0, 2, 7, 17, 37,
77, 157, 317, 637, 1277, 2557, 5117, 10237, 20477, 40957, 81917};
/**
* The upper bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiHigh[SRS_ENTRIES] = {0, 1, 6, 16, 36,
76, 156, 316, 636, 1276, 2556, 5116, 10236, 20476, 40956, 81916, 163836};
The question is: doing that the simulation works, but I found an anomaly.
Setting a value of SRS periodicity = 640, the system works even if we put a
number of UEs greater than this value. For example, I choose 1000 UEs and
SRS periodicity = 640. I would expect this not to work. Does anyone know
why this happens? Is this value of SRS too big, so that it becomes
meaningless?
Thanks in advance
Marco Pennacchioni
--
Posting to this group should follow these guidelines
https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
---
You received this message because you are subscribed to the Google Groups
"ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at https://groups.google.com/group/ns-3-users.
For more options, visit https://groups.google.com/d/optout.

[Jerry Lau]

Hello,

I think we are doing the same research about NB-IoT. how can I get further
contact with you? This topic will be in my Graduation thesis.

Best,
Jerry.


圚 2017幎3月29日星期䞉 UTC+8䞋午9:15:10Marco Pennacchioni写道

Post by 'Marco Pennacchioni' via ns-3-users
Hi everyone!
I'm Marco, I'm developing a master thesis on NB-IoT and LTE systems. I've
a question about the relation beetween the SRS periodicity and the number
of UEs in a LTE simulation. I need to simulate a scenario including 50000
UEs and, reading in a previous thread the problems due to the value of SRS
periodicity, I estimated and implemented SRS values over 640 (note that
*lte-phy.cc , lte-enb-rrc.cc :*
/**
* Sounding Reference Symbol (SRS) periodicity (TSRS) in milliseconds. Taken
* from 3GPP TS 36.213 Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsPeriodicity[SRS_ENTRIES] = {0, 2, 5, 10, 20,
40, 80, 160, 320, 640, 1280, 2560, 5120, 10240, 20480, 40960, 81920};
/**
* The lower bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiLow[SRS_ENTRIES] = {0, 0, 2, 7, 17, 37,
77, 157, 317, 637, 1277, 2557, 5117, 10237, 20477, 40957, 81917};
/**
* The upper bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiHigh[SRS_ENTRIES] = {0, 1, 6, 16, 36,
76, 156, 316, 636, 1276, 2556, 5116, 10236, 20476, 40956, 81916, 163836};
The question is: doing that the simulation works, but I found an anomaly.
Setting a value of SRS periodicity = 640, the system works even if we put a
number of UEs greater than this value. For example, I choose 1000 UEs and
SRS periodicity = 640. I would expect this not to work. Does anyone know
why this happens? Is this value of SRS too big, so that it becomes
meaningless?
Thanks in advance
Marco Pennacchioni

[Nagib Matni]

Hi,
Can you share your code, i'm trying to simulate nb-iot but I can't.

Post by Jerry Lau
Hello,
I think we are doing the same research about NB-IoT. how can I get further
contact with you? This topic will be in my Graduation thesis.
Best,
Jerry.
圚 2017幎3月29日星期䞉 UTC+8䞋午9:15:10Marco Pennacchioni写道

Post by 'Marco Pennacchioni' via ns-3-users
Hi everyone!
I'm Marco, I'm developing a master thesis on NB-IoT and LTE systems. I've
a question about the relation beetween the SRS periodicity and the number
of UEs in a LTE simulation. I need to simulate a scenario including 50000
UEs and, reading in a previous thread the problems due to the value of SRS
periodicity, I estimated and implemented SRS values over 640 (note that
*lte-phy.cc , lte-enb-rrc.cc :*
/**
* Sounding Reference Symbol (SRS) periodicity (TSRS) in milliseconds. Taken
* from 3GPP TS 36.213 Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsPeriodicity[SRS_ENTRIES] = {0, 2, 5, 10, 20,
40, 80, 160, 320, 640, 1280, 2560, 5120, 10240, 20480, 40960, 81920};
/**
* The lower bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiLow[SRS_ENTRIES] = {0, 0, 2, 7, 17,
37, 77, 157, 317, 637, 1277, 2557, 5117, 10237, 20477, 40957, 81917};
/**
* The upper bound (inclusive) of the SRS configuration indices (ISRS) which
* use the corresponding SRS periodicity (TSRS). Taken from 3GPP TS 36.213
* Table 8.2-1. Index starts from 1.
*/
static const uint32_t g_srsCiHigh[SRS_ENTRIES] = {0, 1, 6, 16, 36,
76, 156, 316, 636, 1276, 2556, 5116, 10236, 20476, 40956, 81916, 163836};
The question is: doing that the simulation works, but I found an anomaly.
Setting a value of SRS periodicity = 640, the system works even if we put a
number of UEs greater than this value. For example, I choose 1000 UEs
and SRS periodicity = 640. I would expect this not to work. Does anyone
know why this happens? Is this value of SRS too big, so that it becomes
meaningless?
Thanks in advance
Marco Pennacchioni