Machine Learning Methods for Monitoring of Quasi-Periodic Traffic in Massive IoT Networks
| Autor: | Petar Popovski, Jimmy Jessen Nielsen, René Brandborg Sørensen |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
FOS: Computer and information sciences
network monitoring Computer Science - Machine Learning Computer Networks and Communications Computer science 02 engineering and technology Machine learning computer.software_genre 01 natural sciences Machine Learning (cs.LG) Quality of Service (QoS) Computer Science - Networking and Internet Architecture unevenly spaced time series 0103 physical sciences 0202 electrical engineering electronic engineering information engineering Network performance Cluster analysis 010303 astronomy & astrophysics Networking and Internet Architecture (cs.NI) business.industry Quality of service 020206 networking & telecommunications Network monitoring Unevenly spaced time series Internet of Things (IoT) Computer Science Applications machine learning Hardware and Architecture Lomb-Scargle Signal Processing Artificial intelligence business Wireless sensor network computer Information Systems |
| Zdroj: | IEEE Internet of Things Journal Sørensen, R B, Nielsen, J J & Popovski, P 2020, ' Machine Learning Methods for Monitoring of Quasi-Periodic Traffic in Massive IoT Networks ', IEEE Internet of Things Journal, vol. 7, no. 8, 9046825, pp. 7368-7376 . https://doi.org/10.1109/JIOT.2020.2983217 |
| ISSN: | 2327-4662 |
| DOI: | 10.1109/jiot.2020.2983217 |
| Popis: | One of the central problems in massive Internet-of-Things (IoT) deployments is the monitoring of the status of a massive number of links. The problem is aggravated by the irregularity of the traffic transmitted over the link, as the traffic intermittency can be disguised as a link failure and vice versa. In this article, we present a traffic model for IoT devices running quasiperiodic applications and we present unsupervised, parametric machine learning methods for online monitoring of the network performance of individual devices in IoT deployments with quasiperiodic reporting, such as smart metering, environmental monitoring, and agricultural monitoring. Two clustering methods are based on the Lomb-Scargle periodogram, an approach developed by astronomers for estimating the spectral density of unevenly sampled time series. We present probabilistic performance results for each of the proposed methods based on simulated data and compare the performance to a naïve network monitoring approach. The results show that the proposed methods are more reliable at detecting both hard and soft faults than the naïve-approach, especially, when the network outage is high. Furthermore, we test the methods on real-world data from a smart metering deployment. The methods, in particular the clustering method, are shown to be applicable and useful in a real-world scenario. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|