WiSpec — Commodity Dual-Band Wi-Fi Spectroscopy for Material Classification and Structural Reconnaissance

Author: Abhinav Ranish — Arizona State University Status: Active research project

Open to Collaborators

This is an active research project and we're looking for collaborators — whether you're into wireless systems, signal processing, machine learning, or just curious about RF sensing. If you want to contribute or discuss ideas, reach out: chatgpt@asu.edu

Tactical Applications

WiSpec enables several practical applications across different domains:

• Building Reconnaissance: Using Wi-Fi reflections to characterize interior layouts, wall composition, and structural elements without visual access
• Search and Rescue: Rapid structural assessment to identify building composition, floor materials, and potential hazards during emergency operations
• Smart Building Automation: Material-aware building systems that adapt HVAC, lighting, and comfort systems based on wall and floor composition

Quick Start

Your Hardware

Device	Chipset	Bands	CSI?	Role
Xiaomi Mi Router 4C	MT7628AN (OpenWrt)	2.4 GHz	No	Tier A TX (single-band pilot)
Linux Desktop	Intel AX201 (CNVi)	2.4 + 5 GHz	No (PicoScenes unsupported)	Tier A RX (RSSI dual-band)
Buy: Intel AX200 M.2	Intel AX200	2.4 + 5 GHz	Yes (PicoScenes)	Tier B CSI extraction
Buy: Intel AX210 M.2	Intel AX210	2.4 + 5 + 6 GHz	Yes (PicoScenes)	Tier B+C (adds 6 GHz novelty)

CRITICAL: Your AX201 Cannot Do CSI

The Intel AX201 uses CNVi architecture (MAC in PCH). PicoScenes only supports standalone AX200/AX210. For $15–20, buy an Intel AX200 or AX210 M.2 card + PCIe adapter. This is the single most important purchase.

Your AX201 is still fully usable for RSSI-based dual-band experiments (Tier A).

What You Can Do RIGHT NOW (No New Hardware)

1. Run Tier A RSSI pilot with your Xiaomi router (2.4 GHz, single-band)
2. Run Tier A RSSI dual-band pilot using your AX201 (connect to both 2.4 and 5 GHz networks)
3. Start the FURI proposal

Project Structure

wifi_sensing_research/
├── scripts/
│   ├── tier_a_rssi/
│   │   ├── dual_band_rssi_collector.py    # RSSI logger (laptop side)
│   │   ├── experiment_controller.py       # Experiment orchestrator
│   │   └── openwrt_rssi_logger.sh         # Router-side logger
│   ├── tier_b_csi/
│   │   ├── picoscenes_capture.sh          # CSI capture setup
│   │   └── csi_experiment_controller.py   # CSI experiment orchestrator
│   └── analysis/
│       ├── preprocess_rssi.py             # RSSI data cleaning
│       ├── preprocess_csi.py              # CSI data cleaning
│       ├── feature_extraction.py          # Hand-crafted features
│       ├── classify_materials.py          # ML classifiers + ablation
│       ├── visualize_results.py           # Publication figures
│       └── statistical_tests.py           # Rigorous stats
├── paper/
│   ├── main.tex                           # IEEE workshop paper draft
│   └── references.bib                     # 20+ BibTeX entries
├── funding/
│   ├── furi_proposal.md                   # FURI research proposal
│   ├── personal_statement.md              # FURI personal statement
│   └── timeline.md                        # 14-week timeline
├── data/
│   ├── raw/                               # Raw CSV/CSI files
│   └── processed/                         # Cleaned feature matrices
└── README.md                              # This file

Research Phases

Phase 1: RSSI Pilot (NOW — with current hardware)

# On your Linux desktop:
cd scripts/tier_a_rssi
python3 experiment_controller.py --mode single_band --interface wlan0 --target-ip 192.168.1.1

# On the Xiaomi router (SSH in):
scp openwrt_rssi_logger.sh root@192.168.1.1:/tmp/
ssh root@192.168.1.1 '/tmp/openwrt_rssi_logger.sh /tmp/wifi_stats.csv 1 wlan0'

Phase 2: Dual-Band RSSI (Current hardware, 2 networks)

# Set up your dual-band router with separate SSIDs for 2.4 and 5 GHz
# Then:
python3 experiment_controller.py --mode dual_band \
  --interface wlan0 \
  --ssid-2g "YourNetwork_2G" \
  --ssid-5g "YourNetwork_5G" \
  --target-ip 192.168.1.1

Phase 3: CSI (After buying AX200/AX210)

# Install PicoScenes first (see tier_b_csi/picoscenes_capture.sh for instructions)
cd scripts/tier_b_csi
sudo bash picoscenes_capture.sh --interface wlan1 --channel 6 --bandwidth 20 --duration 60 --output ../data/raw/
python3 csi_experiment_controller.py --interface wlan1 --channel-2g 6 --channel-5g 36

Phase 4: Analysis

cd scripts/analysis
python3 preprocess_rssi.py --input ../../data/raw/ --output ../../data/processed/
python3 feature_extraction.py --input ../../data/processed/ --output ../../data/processed/features.npz
python3 classify_materials.py --input ../../data/processed/features.npz --ablation
python3 visualize_results.py --input ../../data/processed/ --output ../../paper/figures/
python3 statistical_tests.py --input ../../data/processed/ --output ../../paper/tables/

Dependencies

# Python packages
pip install numpy pandas scipy scikit-learn matplotlib seaborn xgboost torch csiread

# System tools (Linux)
sudo apt install iw wireless-tools iputils-ping

Funding

• FURI deadline for Fall 2026: PASSED (was March 18, 2026)
• Next cycle: Spring 2027, deadline ~October 2026
• Proposal is ready in funding/ — personalize and submit
• Alternative: approach a faculty mentor directly with the proposal

Citation and Licensing

License

WiSpec is source-available under a noncommercial license. This is NOT open source (not MIT, not Apache, not GPL). You can read, learn from, and use the code for academic research — but commercial use requires a separate paid license. See LICENSE.md for the full terms and COMMERCIAL-LICENSING.md for commercial inquiries.

Use	Allowed?
Academic research	Yes (with citation)
Personal learning	Yes (with attribution)
Student thesis	Yes (with citation)
Commercial product/service	No — requires paid license
For-profit internal use	No — requires paid license

How to Cite

If you use WiSpec in your research, please cite both the repository and the paper.

Repository:

A. Ranish, "WiSpec: Commodity Dual-Band Wi-Fi Spectroscopy for Material
Classification and Structural Reconnaissance," GitHub, 2026.
https://github.com/abhinav-ranish/WiSpec

Paper (update venue/DOI when published):

A. Ranish, "WiSpec: Commodity Dual-Band Wi-Fi Spectroscopy for Material
Classification and Structural Reconnaissance," [Venue TBD], 2026.

BibTeX:

@software{ranish2026wispec,
  author       = {Ranish, Abhinav},
  title        = {{WiSpec}: Commodity Dual-Band Wi-Fi Spectroscopy for
                  Material Classification and Structural Reconnaissance},
  year         = {2026},
  url          = {https://github.com/abhinav-ranish/WiSpec},
  note         = {Source-available, noncommercial license}
}

@inproceedings{ranish2026wispec_paper,
  author    = {Ranish, Abhinav},
  title     = {{WiSpec}: Commodity Dual-Band Wi-Fi Spectroscopy for
               Material Classification and Structural Reconnaissance},
  booktitle = {[Venue TBD]},
  year      = {2026},
  note      = {Preprint / under review}
}

GitHub also renders a "Cite this repository" button from the CITATION.cff file.

Attribution Requirements

Academic and research users must cite WiSpec in any publication, thesis, report, or presentation that uses this code, methods, or datasets. This is a condition of the license, not just a polite request.

Commercial use of any kind requires prior written permission. Contact: chatgpt@asu.edu

Key References for WiSpec

See paper/references.bib for complete bibliography. Essential reading:

1. Chen et al., "A Survey on Radio Frequency Sensing: From Wi-Fi to 6G," IEEE COMST, 2025
2. Wilson & Patwari, "Propagation Losses in Building Materials: Measurements and Prediction at 915 MHz and 2.4 GHz," IEEE AWPL, 2002
3. Wilson & Patwari, "Radio Tomographic Imaging with Wireless Networks," IEEE TMC, 2010