CIRSense

CIRSense utilizes Channel State Information (CSI) data from Wi-Fi signals for non-contact sensing applications, including distance estimation and respiration rate monitoring.

This project implements the algorithms described in the paper: CIRSense: Rethinking WiFi Sensing with Channel Impulse Response

BibTeX Citation

@ARTICLE{11457720,
  author={Kong, Ruiqi and Chen, He},
  journal={IEEE Transactions on Mobile Computing}, 
  title={CIRSense: Rethinking WiFi Sensing With Channel Impulse Response}, 
  year={2026},
  volume={},
  number={},
  pages={1-15},
  doi={10.1109/TMC.2026.3678326}}

Files Overview

Core Scripts

• cirsense_distance.m: Single-target distance estimation script. Processes CSI data to estimate the distance to a single target and compares it with ground truth.

• cirsense_bpm.m: Single-target respiration rate monitoring script. Estimates breathing rate from CSI data and compares with ground truth respiration data.

• cirsense_distance_multitarget.m: Multi-target distance estimation script.

• cirsense_bpm_multitarget.m: Multi-target respiration rate monitoring script.

Supporting Functions

• build_default_params: Defines default parameters for the algorithms.
• build_default_config: Sets default configuration paths and settings.
• get_window_length: Calculates smoothing window lengths.
• cirsense: Core CIRSense algorithm for distance estimation and respiration trace extraction.
• initialization: Initializes matrices and indices for signal processing.
• Domino: Preprocessing algorithm to correct hardware distortions (e.g., STO, CFO, ...) in CSI data.
• estimate_tau1_freq: Estimates time delays in the frequency domain.

Data

• The input data can be downloaded from https://ieee-dataport.org/documents/cirsense-dataset-real-world-80211ax-csi-measurements-wireless-sensing

Contact

ruiqikong@cuhk.edu.hk

License

This project is for research purposes. Please cite appropriately if used in publications.