13. July 2018 at 17:36 #10855
I’m quite new to this simulator. I’m trying to add some penetration losses for simulating an indoor environment using LTE_config.additional_penetration_loss. However, I’m getting higher average throughput when I introduce additional losses. If I set LTE_config.additional_penetration_loss to ‘indoor’, I get more throughput than with ‘outdoor’, and even more throughput when I set it to ‘deep indoor’. I don’t get it. Am I missing something?
Thanks in advance.
Alberto Alvarez16. July 2018 at 11:39 #10863Fjolla AdemajMember
That all depends on the scenario you are simulating. Setting a user to be indoors, can isolate this user from interference, while the serving base station might be close to the user, so having higher throughput sounds reasonable. However, I will need to know more details about which scenario are you simulating, to give a more detailed answer.
Fjolla16. July 2018 at 13:16 #10864
Thank you very much for your answer Fjolla.
What you say sounds reasonable. However, when having the additional losses set to ‘deep indoor’ (-23dB) I get around 13Mbps average UE throughout. Even though interference is kind of neglected, it seems a little bit high taking into account I have high penetration loss, isn’t it?.
My scenario is a 2×2 MIMO closed-loop spatial multiplexing (tx mode 4), TU channel model, 20MHz BW in a hexagonal grid.
In the end, what I’m trying to simulate is a transmission between an eNB and a UE which is inside a building with very thick walls and a few interference eNBs nearby.
Alberto Alvarez20. July 2018 at 11:31 #10970Fjolla AdemajMember
The throughput results will also depend on the number of users in the cell that share resources, not only on the penetration losses that directly have an impact on the SINR values.
If you consider only 1 base station and one user (omnidirectional antenna and no interference, LTE_config.nr_eNodeB_rings = 0), you will see that increasing the penetration loss will result in lower throughput, as we all would expect. However as soon as you introduce interference, LTE_config.nr_eNodeB_rings = 1, then actually the SINR increases when increasing the penetration losses, due to the isolating effect of interference.
In order to have a better understanding you could use a more realistic scenario. I would suggest to use the 3GPP 3D channel model, where you can place users indoor and outdoor, as well as apply a probability to be in LOS or NLOS. You can change the penetration loss for indoor users in the +macroscopic_pathloss_models/TR36873PathlossModel, by giving different values to parameter wall_pathloss.
Fjolla17. September 2018 at 16:53 #11204
I’ve tested what you suggested to me, but the results still don’t make sense. I ran the 3D channel model example just changing the LTE_config.nr_eNodeB_ring to 0 to have only one base station and I still have a much better performance when I have high penetration losses. I checked the SINR plot values and it’s higher with deep indoor losses than with the outdoor scenario. I’m running the latest version of the simulator (v2.0).
Is there any place in the code where I can modify the CQI mapping table and use 256-QAM modulation?
Thanks again for your help.
- The forum ‘Vienna LTE-A Downlink System Level Simulator’ is closed to new topics and replies.