Open AccessBehavior Detection Mechanism for Trust Sensor Data Using Deep Learning in the Internet of Things
Author(s) -
H Kim,
Eun-Ha Song
Publication year - 2022
Publication title -
webology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.259
H-Index - 18
ISSN - 1735-188X
DOI - 10.14704/web/v19i1/web19301
Subject(s) - computer science , internet of things , network packet , server , wireless sensor network , identification (biology) , deep learning , function (biology) , computer security , real time computing , artificial intelligence , computer network , botany , evolutionary biology , biology
In this paper, we propose BDM-TSD(Behavior Detection Mechanism for Trust Sensing Data) to classify risk group and non-risk group for reliable sensor data identification in IoT environment with sensing function. BDM-TSD collects trust data such as sensing time, operation cycle, and type of sensing data of sensor devices connected to the IoT environment and artificial malicious data. The collected data performs network packet analysis and sensing data behavior analysis through the behavior of the sensor device that is subsequently operated through deep learning. Previously, research was conducted to detect unauthorized system calls of each device through security agents or abnormal behaviors through monitoring servers, and research to detect new and variant malicious behaviors with advanced attack techniques in IoT environments is insufficient A trusted IoT configuration is possible through malicious packet filtering and multi-sensor behavior detection. In this paper, we show how deep learning can be used to detect anomalies and malicious behaviors in the IoT environment based on the sensing function of multiple sensors.