Designing a participatory modelby using ensembleintelligence algorithms to detect security anomalies in wireless sensor-based IoT networks

Abstract

Background: With the development and expansion of technologies and
computing technologies, today the frontiers of knowledge are developing
rapidly. One of the most important factors in starting these technologies is the
formation of a computer network.
purpose: Network traffic is very large and this leads to high data size and
increased noise and makes it very difficult to extract meaningful information to
detect abnormal events. Network training to detect abnormalities helps to
identify the time of the attack. Early detection of attacks improves the stability
of a system.
Methods: The goal of this study is to design a proposed model to identify and
detect malware attacks with appropriate accuracy and reduce the rate of
malfunctions. To meet these goals, we used participatory machine learning
algorithms.
Results: According to the results, our proposed model has a better performance
in intrusion detection of wireless sensor network attacks than the compared
algorithms of decision tree, random tree and Naïve Bayes.
Conclusion: Due to recent advances in data mining, the use of participatory
model in identifying and detecting attacks of wireless sensor networks is very
effective.