Download and Installation

This page describes the steps necessary to install software required for the sensor placement algorithm.

Download

The code can be cloned from the repository

git clone https://github.com/StefanoMarano/SensorPlacement.git

or downloaded as ZIP archive.

Install Python and packages

  • Install Python 3 and the Python package installer by running:

    sudo apt-get install python3 python3-pip

  • The following Python libraries are needed: glob, os, errno, sys, yaml, time, csv, logging, argparse, numpy, scipy, matplotlib, pulp

    On Ubuntu many packages are installed with the default Python 3 installation. You may need to run install them manualy with:

    sudo pip install <package_name>

Install a solver

The computationally intense job is performed by an external, thirt-party, solver. Different solvers can be used including Gurobi, CPLEX, GNU Linear Programming Toolkit. Any solver supported by PuLP can be used.

To install the GNU Linear Programming Toolkit type:

sudo apt-get install glpk-utils

For windows check GLPK for Windows.

For using a different solver, check the documentation of the specific solver.

Legacy MATLAB code with Gurobi

In the folder src_matlab MATLAB scripts for sensor placement are found.

  • Install MATLAB

  • Install and configure the solver Gurobi. You need to get a license to use Gurobi, the license is free for academic use.

  • The code was tested with Gurobi 6.5.1 and Matlab 2013b.

To run the code note the following:

  • The folder ./utils/ needs to be in the path searchable by MATLAB

  • From the MATLAB prompt, configure Gurobi with the script gurobi_setup.m

  • Run the matlab script SensorPlacement.m. Changing some variables in the code allows to modify design parameters including number of sensors and spatial bandwidth. See comments in the file itself.

  • Output is saved in the folder ./output/. Each output MAT file can be loaded into matlab. The variable pos_sol contains the array layout found by the algorithm. The array layout can be plotted with plotArray_C(pos_sol)