mesh-data

Public dataset of Meshcore network scans collected via mesh-utility-tracker.

Data Structure

mesh-data/
├── scans/
│   ├── 2024-01-15/
│   │   ├── batch-1705334400000.csv  ⭐ Browse/search in GitHub
│   │   ├── batch-1705334400000.json
│   │   ├── batch-1705338000000.csv
│   │   └── batch-1705338000000.json
│   └── 2024-01-16/
│       ├── batch-1705420800000.csv
│       └── batch-1705420800000.json
├── deletions/
│   └── 2024-01-15/
│       └── !abcd1234.json
├── index.csv  ⭐ Daily statistics
└── README.md

Data Formats

CSV (Recommended for Filtering)

Click any .csv file in GitHub to browse with built-in table viewer. Perfect for:

• ✅ Searching by radio ID
• ✅ Filtering by signal strength
• ✅ Sorting by any column
• ✅ Quick visual inspection
• ✅ Download filtered results

Columns: `radioId, timestamp, datetime_utc, latitude, longitude, altitude, nodeId, rssi, snr, hopLimit`

Example CSV row:

!abcd1234,1705334400000,2024-01-15T10:00:00.000Z,37.774900,-122.419400,50.0,!def45678,-85,8.5,0

Zero-hop scans: If hopLimit is empty or 0, it's a direct (zero-hop) connection.

JSON (For API Integration)

Each batch file contains an array of scan records:

[
  {
    "radioId": "!abcd1234",
    "timestamp": 1705334400000,
    "location": {
      "lat": 37.7749,
      "lon": -122.4194,
      "altitude": 50
    },
    "nodes": [
      {
        "nodeId": "!def45678",
        "rssi": -85,
        "snr": 8.5,
        "hopLimit": 0
      },
      {
        "nodeId": "!ghi91011",
        "rssi": -92,
        "snr": 5.2,
        "hopLimit": 0
      }
    ]
  }
]

Field Descriptions

• radioId: Unique identifier for the scanning device (Meshtastic node ID)
• timestamp: Unix timestamp in milliseconds (UTC)
• location: GPS coordinates of the scan
  • lat: Latitude in decimal degrees
  • lon: Longitude in decimal degrees
  • altitude: Altitude in meters (optional)
• nodes: Array of detected mesh nodes
  • nodeId: Unique identifier of detected node
  • rssi: Received Signal Strength Indicator (dBm)
  • snr: Signal-to-Noise Ratio (dB)

Data Privacy

User Data Deletion

Users can request deletion of their data by:

1. Opening mesh-utility-tracker Settings
2. Clicking "Delete My Data"
3. Confirming the radio ID
4. Creating a GitHub issue

Deletion records are stored in deletions/YYYY-MM-DD/:

{
  "radioId": "!abcd1234",
  "deletedAt": "2024-01-15T10:30:00.000Z",
  "action": "deletion"
}

Data Anonymization

• Radio IDs are anonymized (only public node identifiers visible)
• No personal information is collected
• GPS coordinates are user-provided and can be obfuscated in the client
• All data is public and contributed voluntarily

Searching & Filtering in GitHub

Browse CSV Files

1. Navigate to scans/YYYY-MM-DD/ folder
2. Click any batch-*.csv file
3. GitHub renders it as a searchable table
4. Use filter box to search columns
5. Click "Raw" → Save to download filtered data

Search All Files

Find scans from specific radio:

!abcd1234 path:scans/ l (CSV)

```python
import pandas as pd
import glob

# Load all CSV files
csv_files = glob.glob('scans/**/*.csv', recursive=True)
dfs = [pd.read_csv(f) for f in csv_files]
all_scans = pd.concat(dfs, ignore_index=True)

print(f"Total scans: {len(all_scans)}")
print(f"Unique radios: {all_scans['radioId'].nunique()}")

# Filter by radio ID
my_scans = all_scans[all_scans['radioId'] == '!abcd1234']

# Filter by signal strength
strong_signals = all_scans[all_scans['rssi'] > -80]

# Export filtered data
my_scans.to_csv('my-filtered-scans.csv', index=False)

Filter with Command Line

# Find all scans from a radio
grep "!abcd1234" scans/**/*.csv


# Find strong signals (rssi > -70)
awk -F',' '$8 > -70' scans/**/*.csv

# Count scans per day
wc -l scans/*/*.csv

# Combine all CSVs for a date range
cat scans/2024-01-{15,16,17}/*.csv > january-week3.csv
### Download Specific Day (CSV)

```bash
curl -O https://raw.githubusercontent.com/just-stuff-tm/mesh-data/main/scans/2024-01-15/batch-1705334400000.csv

Download All Data

git clone https://github.com/just-stuff-tm/mesh-data.git
cd mesh-data

Process with Python

import json
import glob

scan_files = glob.glob('scans/**/*.json', recursive=True)
all_scans = []

for file in scan_files:
    with open(file) as f:
        all_scans.extend(json.load(f))

print(f"Total scans: {len(all_scans)}")
print(f"Unique radios: {len(set(s['radioId'] for s in all_scans))}")

Process with JavaScript

import fs from 'fs';
import path from 'path';

const scanDir = 'scans';
const allScans = [];

function readScans(dir) {
  const files = fs.readdirSync(dir, { withFileTypes: true });
  
  for (const file of files) {
    const fullPath = path.join(dir, file.name);
    if (file.isDirectory()) {
      readScans(fullPath);
    } else if (file.name.endsWith('.json')) {
      const data = JSON.parse(fs.readFileSync(fullPath, 'utf-8'));
      allScans.push(...data);
    }
  }
}

readScans(scanDir);
console.log(`Total scans: ${allScans.length}`);

Data Quality

• Scans are validated before storage
• Duplicate detection based on radioId + timestamp
• Invalid GPS coordinates are rejected
• RSSI/SNR values must be within reasonable ranges

Update Frequency

• New batches committed every 20 scans or 5 minutes
• Batch size may vary based on network activity
• Historical data is never modified (append-only)

License

This dataset is released under CC0 1.0 Universal (Public Domain Dedication).

You are free to:

• Copy, modify, and distribute the data
• Use the data for any purpose, including commercial
• No attribution required (but appreciated!)

Contributing

Data is automatically collected from mesh-utility-tracker users who opt-in to sharing.

To contribute scans:

1. Install mesh-utility-tracker
2. Connect your Meshcore companion device
3. Enable "Share scan data" in Settings
4. Scans are automatically uploaded to this repository

Contact

• Report issues: mesh-utility-tracker/issues
• Request data deletion: connect your radio that you want deleted and Click "Delete My Data"

Statistics

Updated automatically via GitHub Actions

• Total Scans: TBD
• Unique Radios: TBD
• Date Range: TBD
• Last Updated: TBD