ICC2023-NYUSIM-3GPP-NetworkPerformance-Comparison

Title: Full-Stack End-To-End mmWave Simulations Using 3GPP and NYUSIM Channel Model in ns-3
Link to paper: https://arxiv.org/abs/2302.12385

This repository contains the results from ns-3 simulation used in the above mentioned paper and the code files to process the results. If you need assistance with reproducing the results, cofiguring the ns-3 simulation script and in case of any doubts please contact [email protected].

The simulation setup used is as follows: We aim to study the impact of SINR on the end-to-end throughput, latency, and packet drop observed by a wireless modem operating at 28 GHz with a bandwidth of 100 MHz in Urban Microcell (UMi), Urban Macrocell (UMa), Rural Macrocell (RMa), and Indoor Hotspot (InH) scenarios in LOS and NLOS channel conditions using 3GPP Statisitical Channel Model (SCM) and NYUSIM chanel model. Video is transmitted from a remote server using User Datagram Protocol (UDP) at 50 Mbps (4K/8K video, low-end AR/VR applications) over an NR network (ns-3 mmWave module). Simulations performed in this work consider one next-generation base station (gNB) and one user equipment (UE). The UE is fixed at 100 m from the gNB. The gNB transmits at a power of 30 dBm. The simulations are run by fixing the link to a deterministic LOS/NLOS state. We run 50 realizations for each combination of channel model (NYUSIM/3GPP SCM) and channel condition (fixed LOS/NLOS) in each scenario. In each realization (duration of 9 seconds), HARQ is enabled, and blockage is disabled for both 3GPP SCM and NYUSIM.

Output Files

Simulation Run Outputs:- https://drive.google.com/drive/folders/1llWYMZcuhVARXL_5vgzJVHl6c7Yh2_kL?usp=share_link

There are 16 folders each containing 50 simulation runs. The names of the folders are:

1. 3GPPInHLos : 50 simulation runs for InH scenario using 3GPP SCM and channel condition fixed to LOS.
2. 3GPPInHNlos: 50 simulation runs for InH scenario using 3GPP SCM and channel condition fixed to NLOS.
3. 3GPPRmaLos : 50 simulation runs for RMa scenario using 3GPP SCM and channel condition fixed to LOS.
4. 3GPPRmaNlos: 50 simulation runs for RMa scenario using 3GPP SCM and channel condition fixed to NLOS.
5. 3GPPUmaLos : 50 simulation runs for UMa scenario using 3GPP SCM and channel condition fixed to LOS.
6. 3GPPUmaNlos: 50 simulation runs for UMa scenario using 3GPP SCM and channel condition fixed to NLOS.
7. 3GPPUmiLos : 50 simulation runs for UMi scenario using 3GPP SCM and channel condition fixed to LOS.
8. 3GPPUmiNlos: 50 simulation runs for UMi scenario using 3GPP SCM and channel condition fixed to NLOS.
9. nyuInHLos : 50 simulation runs for InH scenario using NYUSIM and channel condition fixed to LOS.
10. nyuInHNlos : 50 simulation runs for InH scenario using NYUSIM and channel condition fixed to NLOS.
11. nyuRmaLos : 50 simulation runs for RMa scenario using NYUSIM and channel condition fixed to LOS.
12. nyuRmaNlos : 50 simulation runs for RMa scenario using NYUSIM and channel condition fixed to NLOS.
13. nyuUmaLos : 50 simulation runs for UMa scenario using NYUSIM and channel condition fixed to LOS.
14. nyuUmaNlos : 50 simulation runs for UMa scenario using NYUSIM and channel condition fixed to NLOS.
15. nyuUmiLos : 50 simulation runs for UMi scenario using NYUSIM and channel condition fixed to LOS.
16. nyuUmiNlos : 50 simulation runs for UMi scenario using NYUSIM and channel condition fixed to NLOS.
    

In each simulation run you have two text files RxPacketTrace.txt and tx-rx-trace.txt.

1. To compute the SINR we use the file RxPacketTrace.txt. We take the average of all SINR values from the column SINR(dB) in the file RxPacketTrace.txt to get the average SINR value over a 9 sec period. Then a CDF of SINR is plotted across all different runs for a particular scenario and channel state. Ex: 3GPPInHLos - in this you have 50 simulation runs. For each run take the average of SINR for 9 second period. This gives you average SINR for the simulation run. Then we plot a ECDF of these SINR values for 50 runs.
2. To compute throughput, latency and packet drops we use the file tx-rx-trace.txt. The file tx-rx-trace.txt contains 5 columns.
   1st column indicates whether the packet is at the Tx or Rx.
   2nd column indicates the timestamp in ns when the packet was sent/received at UDP layer of the Tx/Rx.
   3rd column indicates the size of the packet in bytes.
   4th column indicates the delay in ns.
   5th column indicates the sequence number of the packets.
   

Path Loss Outputs

1 (a). pathlossNYUSIM142GHz.xlsx: has path loss values computed using CI path loss model with 1 m reference distance for UMi, UMa, RMa and InH scenarios.
1 (b). pathlossPlots142GHzNYUSIM.m: plots the path loss values present in the pathlossNYUSIM142GHz.xlsx file.
2 (a). pathlossNYUSIM28GHz.xlsx: has path loss values computed using CI path loss model with 1 m reference distance for UMi, UMa, RMa, InH and InF scenarios.
2 (b). pathLossPlots28GHzNYUSIM.m: plots the path loss values present in thepathlossNYUSIM28GHz.xlsx file.

Code

The code used to process the output files is present in the folder ProcessingCode folder. The ProcessingCode folder contains 7 files:
(i) SystemLevelMetricesCdfs.m - main entry point to the code. Calls methods to compute Latency,Throughput and packet drops.
(ii) ReadTxRxPacketTrace.m - Function to read the tx-rx-trace.txt file contents.
(iii) ReadRxPacketTrace.m - Function to read the RxPacketTrace.txt file contents.
(iv) AverageSINRInEachRun.m - Function to compute the average SINR value for each simulation run.
(v) AverageThroughputInEachRun.m - Function to compute the average throughput value for each simulation run in Mbps.
(vi) AverageLatencyInEachRun.m - Function to compute the average latency value for each simulation run in ms.
(vii) AveragePacketDropsInEachRun.m - Function to compute the average packet drops value for each simulation run in %.