DeepMIMO: A Generic Deep Learning Dataset for  
Millimeter Wave and Massive MIMO Systems
 

Supporting
5G NR Channel Models

DeepMIMO News

What is DeepMIMO?

  • A DeepMIMO Dataset is Completely Defined by (i) the Ray-tracing Scenario and (ii) the Set of Parameters

Accurate Channels

Generated Using the Accurate 3D Ray-tracing Simulator
Wireless InSite by Remcon

Generic and Parameterized

The Key System and Channel Parameters Can be Adjusted for the Target Machine Learning Application

Simple and Reproducible

The DeepMIMO Dataset is Completely Defined by the
Scenario 'R' and the set of Parameters S

Why DeepMIMO?

  • Allowing Dataset Reproducibility for Benchmarking and Comparisons
  • Generating Datasets Compatible with 5G NR Channel Models and Numerologies (in DeepMIMO 5G NR)

How to Generate a DeepMIMO Dataset?

Step 1

Download a DeepMIMO Generator:

Step 2

Download a Ray-Tracing Scenario

Step 3

  • Unzip the scenario folder 
  • Add the unzipped scenario to the DeepMIMO/ray_tracing path
  • Choose the DeepMIMO parameters in DeepMIMO_params.m
  • Run the DeepMIMO generator script following the example

Join the DeepMIMO Community

Sign up for the DeepMIMO mailing list to get the recent updates

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

If you have any question or feedback, please submit it here

License

In order to use the DeepMIMO dataset/scripts or any (modified) part of them, please cite:

1. The DeepMIMO paper: A. Alkhateeb, “DeepMIMO: A Generic Deep Learning Dataset for Millimeter Wave and Massive MIMO Applications,” in Proc. of The Information Theory and Applications Workshop (ITA), San Diego, CA, Feb. 2019. 

@InProceedings{Alkhateeb2019,
author = {Alkhateeb, A.},
title = {{DeepMIMO}: A Generic Deep Learning Dataset for Millimeter Wave and Massive {MIMO} Applications},
booktitle = {Proc. of Information Theory and Applications Workshop (ITA)},
year = {2019},
pages = {1-8},
month = {Feb},
Address = {San Diego, CA}, }

2. The Remcom Wireless InSite website: RemCom, Wireless InSite, “https://www.remcom.com/wireless-insite”. 

@Article{Remcom,

author = {Remcom}, 

title = {{Wireless InSite}}, 

note = {\url{http://www.remcom.com/wireless-insite}.},}

License

In order to use the DeepMIMO dataset/scripts or any (modified) part of them, please cite:

1. The DeepMIMO paper: A. Alkhateeb, “DeepMIMO: A Generic Deep Learning Dataset for Millimeter Wave and Massive MIMO Applications,” in Proc. of The Information Theory and Applications Workshop (ITA), San Diego, CA, Feb. 2019. 

@InProceedings{Alkhateeb2019,
author = {Alkhateeb, A.},
title = {{DeepMIMO}: A Generic Deep Learning Dataset for Millimeter Wave and Massive {MIMO} Applications},
booktitle = {Proc. of Information Theory and Applications Workshop (ITA)},
year = {2019},
pages = {1-8},
month = {Feb},
Address = {San Diego, CA}, }

2. The Remcom Wireless InSite website: RemCom, Wireless InSite, “https://www.remcom.com/wireless-insite”. 

@Article{Remcom,

author = {Remcom}, 

title = {{Wireless InSite}}, 

note = {\url{http://www.remcom.com/wireless-insite}.},}