Particle Filter Demonstration: Illustrates particle filter

This project demonstrates the implementation of a Particle Filter for indoor tracking using simulated RSSI

This project demonstrates the implementation of a Particle Filter for indoor tracking using simulated RSSI (Received Signal Strength Indicator) data. Particle filters are widely used in localization problems, offering robust tracking in environments where noise and uncertainty are prevalent.

Objective

The goal is to track the position of a moving device within a predefined 600x600 grid area by estimating its location using RSSI values from a fingerprint database.

Key Features

1. Fingerprint Database:

• Pre-recorded RSSI values for known (x, y) positions serve as the reference for localization.

1. Particle Filter Framework:

• A set of particles represents possible device locations.
• Each particle is assigned a weight based on the likelihood of its position matching the observed RSSI value.

1. Simulation Process:

• RSSI Simulation: Simulated RSSI measurements at the true device location are perturbed with Gaussian noise to mimic real-world conditions.
• Weight Update: Particle weights are updated using a Gaussian likelihood function comparing predicted and observed RSSI values.
• Resampling: Particles are resampled based on their weights to focus on high-probability areas.
• Motion Update: Motion noise is added to particles to simulate device movement and uncertainty.
• Position Estimation: The estimated position is computed as the weighted average of particle positions.

1. Visualization:

• The simulation dynamically displays particle positions, the true device position, and the estimated position over time, showcasing the convergence of the filter.

Results

The particle filter effectively tracks the device's movement, demonstrating convergence of the estimated position to the true position. The visualization highlights the step-by-step refinement of particle distributions as they adapt to RSSI measurements.

Applications

This simulation is a foundation for practical systems in:

• Indoor Positioning Systems (IPS): Tracking devices in malls, airports, or industrial spaces.
• Wireless Signal Mapping: Optimizing signal strength in networked environments.
• Robotics: Enabling mobile robots to localize in GPS-denied environments.

Conclusion

This work showcases the potential of particle filters for indoor tracking, offering a robust and flexible solution for localization problems. By leveraging RSSI data and particle filtering principles, this simulation provides insights into real-world applications and serves as an excellent learning tool for researchers and practitioners.

You can use and adapt this code for your specific needs, whether for academic purposes or real-world implementation in indoor positioning systems.