The Python code in this repository implements the simulations and plots the figures described in the paper “Spectrum Surveying: Active Radio Map Estimation with Autonomous UAVs” by Raju Shrestha, Daniel Romero, and Sundeep Prabhakar Chepuri.

Python 3.6 or later. Use the package manager pip to install the following packages:

tensorflow 2.4.0
tensorflow-probability 0.12.1 
scipy  1.5.2
cvxpy 1.1.7
matplotlib 3.3.2
pandas 1.2.0 
joblib 1.0.1 
sklearn 0.23.2
opencv-python 4.5.1.48
dill 0.3.3
seaborn 0.11.1 

First, download all the files and folders from this repository. Then, the simulations can be executed by running the file run_experiment.pyfrom the command prompt. One needs to provide the experiment number (e.g. 2001) as an argument while executing the file run_experiment.pyto select the simulation you want to run. The experiments reproducing different figures in the paper are organized in the methods located in the file Experiments/spectrum_surveying_experiments.py. The comments before each method indicate which figure(s) on the paper it generates. For example, to run the experiment no 2001 in the Experiments/spectrum_surveying_experiments.py, in the command prompt, execute the command $ python run_experiment.py 2001.

For more information about the simulation environment, please check here.

If our code is helpful in your research or work, please cite our paper: “Spectrum Surveying: Active Radio Map Estimation with Autonomous UAVs.”

Please feel free to contact us by email if you have any issues in running the code.