DAPS -- A web-based system for sensor data prediction

Ignov, Ljubomir and Arsov, Jordan (2017) DAPS -- A web-based system for sensor data prediction. In: PROCEEDINGS of the 14th Conference on Informatics and Information Technology. Faculty of Computer Science and Engineering, Ss. Cyril and Methodius University in Skopje, Macedonia, Skopje, Macedonia, pp. 185-189. ISBN 978-608-4699-07-1

[img]
Preview
Text
978-608-4699-07-1_pp185-189.pdf

Download (490kB) | Preview
Official URL: http://ciit.finki.ukim.mk

Abstract

Since raw sensor measurements are voluminous and require high bandwidths to be sent to the data centers, data prediction is one useful approach for data reduction. Various algorithms based on different models for times series prediction can be used for this purpose. In this paper, we will present the development of a DAta Prediction System (DAPS), a web-based system that predicts future sensor measurements. Our online tool performs data prediction on one-dimensional and multidimensional sensor readings, using 3 different algorithms: Least Mean Square (LMS), Least Mean Square with variable step size (LMS-VSS), and Moving Average (MA) of different orders. Additionally, the visualization engine from DAPS visualizes the results obtained from the data prediction, by means of MSE and percentage of data reduction. This tool can also compare the performances provided by the three algorithms, since, depending on the nature of the sensor data, algorithms perform differently for different sensor measurements. We believe that our web-based system for sensor data prediction can be utilized by future developers of wireless sensor networks (WSN) and Internet of things (IoT) developers, who can choose the best technique for reducing sensor measurements.

Item Type: Book Section
Subjects: International Conference on Informatics and Information Technologies > Students Session
Depositing User: Vangel Ajanovski
Date Deposited: 29 Nov 2017 18:28
Last Modified: 29 Nov 2017 18:28
URI: http://eprints.finki.ukim.mk/id/eprint/11401

Actions (login required)

View Item View Item