WiShield

Introduction

This project encompasses the data and processing codes utilized in the Evaluation section of the paper "WiShield: Fine-grained Countermeasure Against Malicious Wi-Fi Sensing in Smart Home". The data for each test is organized within the "Data/" folder, while the associated processing scripts are put in the "Figure/" folder. Within the "Figure/" folder, each ".mlx" file delineates the methodology for data processing and figure generation. Below is the detailed file tree for our project.

Project Structure

A description of the project's directory structure, including key files and directories.

WiShield/
│
├── Basic5300Tools/         		# Tools to decode the collected CSI files (provided by [1])
├── My5300Tools/     			# Customized tools to preprocess CSI data
│   └── GetNewACSIVarienceIndicator.m	# Calculation of amplitude variance indicator for motion detection
│   └── ......
├── spotfi/   	                        # SpotFi implementation (modified from [2])
├── Data/				# Collected CSI data and processing results
│   └── AoAEstimation/		        # Data and results for evaluating obfuscation of AoA estimation
|       ├── calibration/		# Data for internal phase offset calibration
|       ├── move_test/			# Evaluation data with moving target
|       ├── still_test/			# Evaluation data with still target
│   └── MotionDetection/		# Data and results for evaluating obfuscation of motion detection
|       ├── DRandFDR/		        # Data and results for evaluating obfuscation of detection of temporary motion
|       ├── SpatialDistribution/	# Data and results for evaluating obfuscation of detection of long-lasting motion
├── Figure/				# Code for generating Figure 10-11 and 13-15
└── README.md           		# Project README

[1] https://dhalperi.github.io/linux-80211n-csitool/

[2] https://bitbucket.org/mkotaru/spotfimusicaoaestimation/src/master/

Tested Environment

List all the hardware and software tested for running the project.

• Hardware: AMD Ryzen 7 4700U

• Software: Windows 10, MATLAB R2023b

How to use

I. Prepare MATLAB

1. You can download and install MATLAB by following the instructions at this URL: https://www.mathworks.com/help/install/ug/install-products-with-internet-connection.html

2. Install the following MATLAB add-ons by following the instructions at this URL: https://www.mathworks.com/help/matlab/matlab_env/get-add-ons.html

   • Signal Processing Toolbox
   • Wavelet Toolbox
   • Statistics and Machine Learning Toolbox
   • Parallel Computing Toolbox (Optional, for parallel calculation of AoA estimation)

II. Download WiShield

You can obtain WiShield with git

git clone https://github.com/zlab-pub/WiShield.git

or download ZIP.

III. Run the Codes

In Figure/ folder, the files are the live scripts containing codes for processing CSI data and generating figures. They can be run through the following steps.

1. Change the current folder to WiShield/ in MATLAB.
2. Open the ".mlx" files in Figure/ without changing the current folder, and then, they can be run directly.

In each ".mlx" file, the first block adds the paths of the dependent libraries. They are relative paths, so the current folder in MATLAB should be WiShield/ to import the libraries successfully.