Internet of Things (IoT) devices are starting to play a predominant role in our everyday life. Application systems like Amazon Echo and Google Home allow IoT devices to answer human requests, or trigger some alarms and perform suitable actions. In this scenario, any data information, related device and human interaction are stored in databases and can be used for future analysis and improve the system functionality. Also, IoT information related to the network level (wireless or wired) may be stored in databases and can be processed to improve the technology operation and to detect network anomalies. Acquired data can be also used for profiling operation, in order to group devices according to their characteristics. LoRaWAN (Long Range Wide Area Network) is one of the emerging IoT technologies in today’s world, it is a protocol based on LoRa modulation. In this work, we propose a methodology to process LoRaWAN packets and perform profiling of the IoT devices. Specifically, we use the k-means algorithm to group devices according to their radio and network behaviour. We tested our approach on a real LoRaWAN network where the entire captured traffic is stored in a proprietary database. Our analysis, performed on 286, 753 packets with 765 devices involved, leads to remarkable clustering performance according to validation indices such as the Silhouette and the Davies-Bouldin indices. Further, with the help of field-experts, we were able to analyze clusters’ contents, revealing results both in line with the current network behaviour and alerts on malfunctioning devices, remarking the reliability of the proposed approach.

Valtorta J.M., Martino A., Cuomo F., Garlisi D. (2019). A clustering approach for profiling LoRaWAN IOT devices. In B.D.R. Ioannis Chatzigiannakis (a cura di), Ambient Intelligence 15th European Conference, AmI 2019, Rome, Italy, November 13–15, 2019, Proceedings (pp. 58-74). Springer [10.1007/978-3-030-34255-5_5].

A clustering approach for profiling LoRaWAN IOT devices

Garlisi D.
2019-01-01

Abstract

Internet of Things (IoT) devices are starting to play a predominant role in our everyday life. Application systems like Amazon Echo and Google Home allow IoT devices to answer human requests, or trigger some alarms and perform suitable actions. In this scenario, any data information, related device and human interaction are stored in databases and can be used for future analysis and improve the system functionality. Also, IoT information related to the network level (wireless or wired) may be stored in databases and can be processed to improve the technology operation and to detect network anomalies. Acquired data can be also used for profiling operation, in order to group devices according to their characteristics. LoRaWAN (Long Range Wide Area Network) is one of the emerging IoT technologies in today’s world, it is a protocol based on LoRa modulation. In this work, we propose a methodology to process LoRaWAN packets and perform profiling of the IoT devices. Specifically, we use the k-means algorithm to group devices according to their radio and network behaviour. We tested our approach on a real LoRaWAN network where the entire captured traffic is stored in a proprietary database. Our analysis, performed on 286, 753 packets with 765 devices involved, leads to remarkable clustering performance according to validation indices such as the Silhouette and the Davies-Bouldin indices. Further, with the help of field-experts, we were able to analyze clusters’ contents, revealing results both in line with the current network behaviour and alerts on malfunctioning devices, remarking the reliability of the proposed approach.
2019
Settore ING-INF/03 - Telecomunicazioni
Valtorta J.M., Martino A., Cuomo F., Garlisi D. (2019). A clustering approach for profiling LoRaWAN IOT devices. In B.D.R. Ioannis Chatzigiannakis (a cura di), Ambient Intelligence 15th European Conference, AmI 2019, Rome, Italy, November 13–15, 2019, Proceedings (pp. 58-74). Springer [10.1007/978-3-030-34255-5_5].
File in questo prodotto:
File Dimensione Formato  
Valtorta2019_Chapter_AClusteringApproachForProfilin.pdf

Solo gestori archvio

Tipologia: Versione Editoriale
Dimensione 1.23 MB
Formato Adobe PDF
1.23 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10447/439303
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 13
  • ???jsp.display-item.citation.isi??? 11
social impact