This repository is a complement to the paper Budgeted sensor placement for source localization on trees [1] presented at LAGOS conference in May 2015. It contains a Python implementation of the algorithms described in the paper, together with testing scripts.
Also, the repository contains the extended version of [1].
For any question, suggestion or comment please write to Brunella Spinelli.
- sensor_placement/prob_err.py
algorithm to optimally allocate k sensors in order to minimize the error probability in source localization, i.e., the probability of obtaining an estimated source different from the actual source of the diffusion
- sensor_placement/exp_dist.py
algorithm to optimally allocate k sensors in order to minimize (in expectation) the distance between the estimated source and the actual one
- test_prob_err.py, test_exp_dist.py
scripts to test the above algorithms on randomly generated trees
This implementation requires Python 2.7 and the following third-party libraries: NetworkX, functools32.
[1] L. E. Celis, F. Pavetić, B. Spinelli, P. Thiran, Budgeted Sensor Placement for Source Localization on Trees, LAGOS 2015