Energy saving in WSN using monitoring values predictionReport as inadecuate

Energy saving in WSN using monitoring values prediction - Download this document for free, or read online. Document in PDF available to download.

1 LAAS-SARA - Équipe Services et Architectures pour Réseaux Avancés LAAS - Laboratoire d-analyse et d-architecture des systèmes Toulouse 2 REDCAD - Unité de Recherche en développement et contrôle d-applications distribuées 3 LAAS - Laboratoire d-analyse et d-architecture des systèmes Toulouse

Abstract : The Wireless Sensor Networks WSNs deployment introduces many issues and challenges mainly in terms of energy independence. In this context, we adopted the IBM control loop which is composed of four steps Monitor, Analyze, Plan and Execute to manage Quality of Service QoS 1. This paper focuses on the first step which consists in monitoring and sending periodically QoS values such as the value of power remaining in the battery of each sensor. We notice that the transmission process is very costly in terms of energy and reduces the battery lifetime. In this work, we propose a probabilistic approach that estimates a part of these QoS monitoring values and therefore economizes their transmission energy and extends the sensor battery lifetime. Our approach is based on the hidden Markov chain and the fuzzy logic. It is composed of two steps: i learning which allows apprehending the WSNs behavior and ii prediction which estimates QoS monitoring values. A WSN application deployed in a datacenter is studied as an illustration. The carried out experiments over AZEM WSN simulator show that the gain varies from 25% to 75% of the battery energy.

Keywords : Wireless sensor network probabilistic approach hidden Markov chain monitoring prediction

Author: Ahmed Jemal - Marwa Hachicha - Riadh Ben Halima - Ahmed Hadj Kacem - Khalil Drira - Mohamed Jmaiel -



Related documents