CoLoNet: Collaborative Localization in Wireless Sensor Networks

CoLoNet simulates how mobile wireless sensors can localize themselves more accurately by talking to each other. Using geometric uncertainty models and neighbor-based trilateration, it shows how inter-tag links can reduce localization error—especially when fixed anchors are few and far between.

Project Structure

colonet/ ├── assets/ ├── results/ ├── main.py # Core simulation and visualization ├── report.pdf # Technical documentation ├── presentation.pdf # Presentation slides ├── README.md ├── LICENSE └── .gitignore

Features

• Mobile + fixed sensor modeling with configurable sensing radii
• Real-time uncertainty visualization using matplotlib
• Neighbor-based updates with optional tag-to-tag communication
• Custom coordinate and sensor classes with basic physics support
• Simulation toggle for enabling/disabling tag links

Getting Started

Dependencies

This project requires Python 3.9+ and the following packages:

bash pip install numpy matplotlib pillow imageio[ffmpeg]

Run the Simulation

bash python main.py

This will launch a fullscreen live simulation showing: - Fixed anchors (colored ×) - Mobile tags (colored circles) - Dynamic links (anchor–tag, tag–tag) - Marker size ∝ uncertainty

Output Snapshots

| Without Tag Links | With Tag Links | |-------------------|----------------| | | |

The presence of inter-tag communication prevents uncertainty divergence over time.

Sample Output Video

Each simulation produces a sim.mp4 video showing localization evolution over time.

| Mode | Example Video | |-------------------|-------------------------| | No Tag Comm | results/notagcomm/sim.mp4 | | With Tag Comm | results/tagcomm/sim.mp4 |

Configuration

Edit the following parameters directly in main.py:

python tagcomm = True # Enable/disable tag-to-tag communication final_time = 10 # Total simulation time in seconds sensing_radius = 7 # Sensor range in meters timestep = 0.1 # Timestep in seconds

Applications

• Wildlife localization and monitoring
• Ad hoc sensor networks in urban environments
• Low-power distributed positioning systems
• Indoor GPS-denied navigation

Project Context

Developed as part of the course project for EE 617: Sensors in Instrumentation at IIT Bombay.

Author:
Aaron John Sabu
Department of Electrical Engineering
Indian Institute of Technology Bombay

License

MIT License © 2020
Indian Institute of Technology Bombay