Introduction

This is the implementation of this project. It contains two folders, namely differentiation and impuation, which corresponds to the two parts of paper respectively.

Requirements

• Pytorch 1.8.1
• Numpy 1.19.2
• Pandas 1.1.3
• Sklearn 0.24.1
• Matplotlib 3.3.2
• Shapely 1.8.1

You may use " pip3 install -r requirements.txt" to install the above libraries.

Usage

Step 1: differentiate MARs and MNARs

cd ./differentiation ; python3 differentiator.py --site KDM --method DasaKM --thre 0.1

after this step, a csv file with differentiated results will be generated in the data folder.

Step 2: generate a json file for the input of BiSIM

cd ../imputation/preprocess ; python3 generate_input_json.py --site KDM --method DasaKM --thre 0.1 

An input file in json format will be generated in the data folder.

Step 3: run BiSIM model for imputation

cd ../ ; nohup python3 -u main.py --site KDM --method DasaKM --thre 0.1 --epochs  500 --batch_size 32 > results.txt & 

Explaination of Parameters

site: the building, e.g., KDM or WDS.

method: the differentiator, e.g., DasaKM or TopoAC.

thre: the in-cluster differentiation threshold, e.g., thre=0, 0.1, 0.2.

batch_size: the number of samples for back propagation in one pass

epochs: the number of training rounds

Acknowledgements

We appreciate the work of BRITS and SSGAN, and their contributed codes available in here for BRITS and here for SSGAN.

Citation

Please cite our paper if you use it for research purposes.

@inproceedings{li2023data,
  title={Data Imputation for Sparse Radio Maps in Indoor Positioning},
  author={Li, Xiao and Li, Huan and Chan, Harry Kai-Ho and Lu, Hua and Jensen, Christian S},
  booktitle={2023 IEEE 39th International Conference on Data Engineering (ICDE)},
  pages={2235--2248},
  year={2023},
  organization={IEEE}
}