Vienna 5G System Level Simulator in MATLAB — Full Walkthrough & Transmit Power Study

A complete walkthrough of the Vienna 5G System Level Simulator (5GSL) in MATLAB — what system-level simulation actually is, how the simulator executes, and a controlled experiment on base-station transmit power with results that were not what I expected. This was produced for EEE 4854 (Cellular Communication Laboratory) at the Islamic University of Technology. ━━━━━━━━━━━━━━━━━━━━━━ WHAT'S IN IT ━━━━━━━━━━━━━━━━━━━━━━ The scenario is scenarios.basicScenario, shipped with the simulator and left unmodified: one omnidirectional 4x4 macro base station serving one indoor 2x2 user in a 500 m x 300 m region, at 2 GHz over 20 MHz, free-space path loss, PedA channel, round-robin scheduling, LTE downlink transmit mode 4. The controlled variable is the base-station antenna transmit power — swept over eight values from the shipped 40 W (46 dBm) down to 4 microwatt (-24 dBm). Experiment A uses the shipped topology. Experiment B repeats the identical sweep with one added identical macro base station 170 m away, which turns the link from noise-limited into interference-limited. Four results: 1. Experiment A is exactly noise-limited. Wideband SINR falls 10.0 dB for every 10 dB of transmit power removed — 68.44 dB down to -1.56 dB. That matches a link-budget calculation of ~68 dB done by hand before running anything. 2. Throughput is completely flat at 159.74 Mbit/s from 46 dBm all the way down to 16 dBm, while the SINR behind it drops 30 dB. That plateau belongs to the link performance model, not the link quality model — the CQI table has simply saturated. 3. Experiment B pins at 13.11 dB and will not move, however much power you add. The predicted signal-to-interference ratio from geometry alone was 13.1 dB, and the per-antenna receive powers confirm it independently at 13.111 dB. 4. The one I did not predict: with the interferer present, the reported rank indicator stays near 2 while the channel only supports one stream. Link adaptation over-reaches, mean BLER hits 0.89, and median user throughput falls to ZERO at a perfectly healthy 13 dB SINR. ━━━━━━━━━━━━━━━━━━━━━━ TWO TRAPS WORTH KNOWING ━━━━━━━━━━━━━━━━━━━━━━ If you use this simulator yourself, these two cost me time: antenna.transmitPower is in WATT, not dBm. 40 in the scenario file means 40 W = 46 dBm. Reading it as dBm is a 6 dB error and nothing warns you. widebandSinr and effectiveSinr come out of the result object ALREADY in dB. Only SNR_DL and SINR_DL are linear. Converting the first two again silently gives plausible-looking nonsense. ━━━━━━━━━━━━━━━━━━━━━━ TOOLS ━━━━━━━━━━━━━━━━━━━━━━ Vienna 5G System Level Simulator v1.1 — Institute of Telecommunications, TU Wien (academic licence). MATLAB R2024a. Simulator homepage: https://www.tuwien.at/etit/tc/en/vien... Reference paper: M. K. Müller et al., "Flexible multi-node simulation of cellular mobile communications: the Vienna 5G System Level Simulator", EURASIP Journal on Wireless Communications and Networking, 2018:227. #5G #MATLAB #WirelessCommunications #NetworkSimulation #Telecommunications #SINR #ViennaSimulators