LENA: Throughput vs Distance

1,431 views
Skip to first unread message

Longhao Zou

unread,
Sep 28, 2012, 9:49:31 AM9/28/12
to ns-3-...@googlegroups.com
Dear all,

I have updated the LENA to the latest version and modified "lena-simple-epc.cc". I set up a simple scenario just like this:
1. 1 UE attached to 1 eNodeB
2. RB=50 (BW=10MHz, Maximun throughput=33.53Mbps); Txpower=10 dBm; DlEarFcn=100;
3. UE is moved away from the position of the eNodeB to 500m

However, I got some unexpected results of throughput of that UE shown as below. In the figure, you can see that throughput equals to 0 at about 100m. Could anyone tell me what happened to the connection between UE and eNodeB at about 100m, please? That has confused me since several weeks ago. Many thanks to anyone can help in advance!

Best regards,
Longhao
1-1-500.jpg
mcs.jpg

Longhao Zou

unread,
Sep 28, 2012, 10:39:25 AM9/28/12
to ns-3-...@googlegroups.com
Hello again,

This is my scenario script attached to this email.

regards,
Longhao

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/KKce6eeI1IcJ.
To post to this group, send email to ns-3-...@googlegroups.com.
To unsubscribe from this group, send email to ns-3-users+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ns-3-users?hl=en.

lteepc-moving-ue-4.cc

Jordan Mihaylov

unread,
Oct 2, 2012, 7:17:40 AM10/2/12
to ns-3-...@googlegroups.com
 Hi,

consider boosting the "ns3::LteUePhy::TxPower". In my experimets I used 23 dBm (the max permited tx power for LTE-800MHz UEs in Europe). The default value was 10 dBm but I guess that's ok only for  relatively small distances to the eNB attached to.

Cheers
Jordan

Longhao Zou

unread,
Oct 2, 2012, 10:27:04 AM10/2/12
to ns-3-...@googlegroups.com
Thanks to Jordan, I will have a try.

Regards,
Longhao

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/ncwIXwj4TewJ.

Nicola Baldo

unread,
Oct 2, 2012, 3:44:12 PM10/2/12
to ns-3-...@googlegroups.com
Hi Longhao,

 looking at your MCS figure, I don't think that the problem is in the power, I mean, even if you change the power I would suspect that the problem might still appear for a different value of the distance.
If you could post RlcStats and PdcpStats maybe we could make a better guess at where the problem is.
Also, can you reproduce the same behavior in a static scenario? I mean, if you place the UE statically at 99m at the beginning of the simulation, do you still get 0 throughput?

Nicola

Longhao Zou

unread,
Oct 2, 2012, 4:35:37 PM10/2/12
to ns-3-...@googlegroups.com

Hi Nicola,

Thanks to your prompt!
I haven't tried to change the txpower of LteUephy but I have tried to changed the txpower of LteEnbphy. When I used 20dBm foAndr LteEnbPhy, that breakpoint appearred at about 300m. So I am not sure if that would be the fault of txpower in LteUephy.
And I also uploaded the trace files of the scenario in my first email please find them. Next I will try the static scenario and let you know the result asap. Thank you!

regards,
Longhao

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/oj1QX41gggwJ.
DlPdcpStats.txt
DlRlcStats.txt

Longhao Zou

unread,
Oct 2, 2012, 4:40:11 PM10/2/12
to ns-3-...@googlegroups.com
Hello again,

I forgot to tell you that the UE moves away from eNodeB in 1 meter every 10 seconds. So do you think about that those breakpoints would be caused by the fading? Because I didn't consider the fading model...

Regards,
Longhao

Longhao Zou

unread,
Oct 3, 2012, 11:52:15 AM10/3/12
to ns-3-...@googlegroups.com
Hi Nicola,

I received 0 packet when I placed the UE at 97m (because the throughput of the points from 95m to 99m was 0 in the scenario mentioned in my first email). And MCS was OK. I have uploaded the DlRlcStats.txt and DlPdcpStats.txt in this email.
Hopefully that would help to solve this problem. Thanks a lot!

Regards,
Longhao

On 2 October 2012 20:44, Nicola Baldo <nba...@cttc.es> wrote:

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/oj1QX41gggwJ.
DlMacStats.txt
DlPdcpStats.txt
DlRlcStats.txt

Nicola Baldo

unread,
Oct 4, 2012, 8:14:03 AM10/4/12
to ns-3-...@googlegroups.com
Thanks Longhao for the detailed information. Looking at MacStats and RlcStats, I suspect it could be due to a bug in the AMC. I've just submitted bug 1508 to track this issue.

Longhao Zou

unread,
Oct 4, 2012, 11:03:00 AM10/4/12
to ns-3-...@googlegroups.com
Thanks to Nicola!

On 4 October 2012 13:14, Nicola Baldo <nba...@cttc.es> wrote:
Thanks Longhao for the detailed information. Looking at MacStats and RlcStats, I suspect it could be due to a bug in the AMC. I've just submitted bug 1508 to track this issue.

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/H5TUu7fSVaoJ.

Marco

unread,
Oct 9, 2012, 8:36:24 AM10/9/12
to ns-3-...@googlegroups.com
Hi,

I tried o work a bit on the AMC and I experienced a different behavior.

I've tried to run the simulation script for distances equal to 97, 98 and 99, and in the simulation the error rate is mostly 0. I used the same simulation script, I've just modified the first position in order to speed up the test (I do not have with me a machine with enough performance for running the whole scenario), in attachment the script I used.


Best regards,

marco.

lteepc-moving-ue-4-97m.cc

Longhao Zou

unread,
Oct 9, 2012, 9:34:41 AM10/9/12
to ns-3-...@googlegroups.com
Hi Marco,

I am wondering there is some problem in the LTE AMC model. Now the script runs very well if I am using the PiroEW2010 model by adding "Config::SetDefault ("ns3::LteAmc::AmcModel", EnumValue (LteAmc::PiroEW2010));".
So you can do some test on the MiErrorModel, the default AMC model...

Regards,
Longhao

To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/Rdw7BOMo93UJ.

Marco Miozzo

unread,
Oct 9, 2012, 9:45:47 AM10/9/12
to ns-3-...@googlegroups.com
Hi Longhao,

the test I reported in the previous email works with MiErrorModel and
returns an error rate at MAC, RLC and PDCP almost null; therefore I
would exclude that is problem of the AMC.

My two cents,
marco.

Longhao Zou

unread,
Oct 9, 2012, 9:58:54 AM10/9/12
to ns-3-...@googlegroups.com
Hi Macro,

My apologies.
Then it still looks so strange...I upload two figure about the throughput and MCS versus distance by using the PiroEW2010...Hopefully it would benefit for you...

Regards,
Longhao
mcs.jpg
throughput.jpg

Nicola Baldo

unread,
Oct 17, 2012, 11:13:34 AM10/17/12
to ns-3-...@googlegroups.com
Hi Longhao,

as reported by Marco, we could not reproduce the bug in a static scenario based on your program. Hence the bug was marked as invalid. If you can generate a program that can successfully reproduce the bug in a static scenario, we could reopen the bug (in this case I would appreciate very much if we could follow up the discussion on bugzilla).

Regards,

Nicola


Longhao Zou

unread,
Oct 17, 2012, 11:18:28 AM10/17/12
to ns-3-...@googlegroups.com
Hi Nicola,

Thanks to you. I will let you know if I generate another useful scenario.

Regards,
Longhao

--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ns-3-users/-/pqU29mV71BAJ.
Reply all
Reply to author
Forward
0 new messages