This repository contains code supporting "SMILE: Robust Network Localization via Sparse and Low-Rank Matrix Decomposition". For any questions, please reach out to Lilly Clark ([email protected]) or Sampad Mohanty ([email protected]).
To test SMILE, our novel approach to network localization, on a simulated dataset of 500 nodes with 50 anchors, run smile.py
.
To test generating simulated data, run process_data.py
.
To test a graph convoluntional network for large-scale network localization, run gcn.py
.
For more details, see https://github.com/Yanzongzi/GNN-For-localization.
To test a baseline approach of rank reduction via PSVD and embedding via multidimensional scaling, run mds.py
.
To compare each of these approaches on a network of 500 nodes and 50 anchors, run run_test.py
.
Requirements
python 3.6.9
torch 1.10.2
torch-geometric 2.0.3
numpy 1.18.5