Compressed sensing with model based reconstruction

Imrich Andráš, Ján Šaliga, Linus Michaeli

Abstract

This work presents a way of reducing the power consumption of the existing wireless sensor network used for monitoring of water quality. The improvement of sensor nodes energy efficiency is achieved by utilizing compressed sensing and lowering the average sampling rate. A novel continuous model based reconstruction method is proposed, utilizing a water parameter signal model rather than a discrete dictionary. No trained dictionary and hence no signal database is required a priori to compressed sensing application.

Keywords

compressed sensing, sparse signal, nonuniform sampling, model based reconstruction, water quality monitoring

Download

IMEKO-TC4-2019-050.pdf

DOI

-

IMEKO TC

TC4 - Measurement of Electrical Quantities

Back to the proceedings