tangle-network-particle-filter

By Or Tslil, Tal Feiner and Avishy Carmi This package is part of the work titled "Distributed Information Fusion in Tangle Networks".

Algorithm

The package contains a fusion scheme for distributed particle filters and a demo for sensor network. The fusion scheme is based on the generalization of the Covariance Intersection algorithm. The fusion algorithm is fed with a network transition matrix that contains the connectivity of the network (see demo.py).

Demo

The demo implemented here is a sensor network for object positioning using range and bearing observations. The target stochastic model is of a sircular walk, and the estimated state is the target center of mass and orientation. An example for the target trajectory and a position of the sensor is shown in the next figure.

For this demo we use a random network transition matrix as follows.

Installation

1. Clone the repository:

git clone https://github.com/ortslil64/tangle-network-particle-filter.git

2. Install the dependencies in requirments.txt file

3. Install the package using pip:

pip install -e tangle-network-particle-filter

Usage

Run the demo file:

python3 demo/demo.py  

API tutorial

TO DO

References

[1] Tslil O, Feiner T, Carmi A. Distributed information fusion in tangle networks. Automatica.;125:109417.

[2] J. Li and A. Nehorai. Distributed particle filtering via optimal fusionof gaussian mixtures.IEEE Transactions on Signal and InformationProcessing over Networks, 4(2):280–292, 2018.