An Energy Efficient Integration Model for Sensor Cloud Systems

Ngoc-Thanh Dinh

doi:10.1109/ACCESS.2018.2886806

Return

An Energy Efficient Integration Model for Sensor Cloud Systems

2019-01-01

Ngoc-Thanh Dinh

Younghan Kim *

DOI：10.1109/ACCESS.2018.2886806

Original

Save

View PDF

Abstract

En 中文

This paper proposes a novel information-centric prediction-based integration model for WSNs and the sensor cloud, which exploits the trade-off between the data accuracy requirement of applications, sensing data prediction quality, and the energy efficiency of sensors, to reduce the workloads and energy consumptions of resource-constrained sensors. The model decouples information producers (IPDs) (i.e., physical sensors) from information providers (IPVs), which are implemented as IPDs' virtual sensors in the sensor cloud, to enable IPVs to provide sensing services when IPDs sleep. In the model, we design an efficient interactive sensing data prediction scheme for IPDs and IPVs to predict and control the accuracy of the sensing data prediction of IPVs using internal temporal information correlation, community detection, and external information correlation among sensors. According to the data accuracy requirement of applications, the model controls: 1) the number of IPDs required to be active and 2) when an active IPD transmits sensing data to the sensor cloud, to maintain the quality of sensing data meeting the requirement. Through extensive experiments with data collected from the real-world IntelLab sensor deployment, we show that the model achieves significant improvements in terms of data transmission suppression ratio, energy efficiency, and response latency compared with the existing schemes.

Keywords:

Information centric wireless sensor networking

energy efficiency

sensor cloud

IoT cloud

interactive sensor data prediction

machine learning

sensor quality of information

SDN/NFV

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.