The interest in the Internet of Things (IoT) is increasing both as for research and market perspectives. Worldwide, we are witnessing the deployment of several IoT networks for different applications, spanning from home automation to smart cities. The majority of these IoT deployments were quickly set up with the aim of providing connectivity without deeply engineering the infrastructure to optimize the network efficiency and scalability. The interest is now moving towards the analysis of the behavior of such systems in order to characterize and improve their functionality. In these IoT systems, many data related to device and human interactions are stored in databases, as well as IoT information related to the network level (wireless or wired) is gathered by the network operators. In this paper, we provide a systematic approach to process network data gathered from a wide area IoT wireless platform based on LoRaWAN (Long Range Wide Area Network). Our study can be used for profiling IoT devices, in order to group them according to their characteristics, as well as detecting network anomalies. Specifically, we use the k-means algorithm to group LoRaWAN packets according to their radio and network behavior. We tested our approach on a real LoRaWAN network where the entire captured traffic is stored in a proprietary database. Quite important is the fact that LoRaWAN captures, via the wireless interface, packets of multiple operators. Indeed our analysis was performed on 997, 183 packets with 2169 devices involved and only a subset of them were known by the considered operator, meaning that an operator cannot control the whole behavior of the system but on the contrary has to observe it. We were able to analyze clusters’ contents, revealing results both in line with the current network behavior and alerts on malfunctioning devices, remarking the reliability of the proposed approach.
|Numero di pagine||15|
|Rivista||Journal of Ambient Intelligence and Humanized Computing|
|Stato di pubblicazione||Published - 2021|
All Science Journal Classification (ASJC) codes