ABSTRACT
In an era of the industrial internet of things (IoT), data transferred or saved is always vulnerable to attacks.
The IoT networks are needed for implementing security in IoT devices. The IoT networks are considered as
secured with authentication and encryption, but these networks are not protected against cyber-attacks.
Although there exist hundreds of data protection systems, but there are some shortcomings as well. Thus,
anomaly detection takes the responsibility upon itself to make various kinds of attacks less vulnerable.
This is achieved by making use of the power of data mining algorithms and tools to analyze and capture
any anomalous network traffic. Swarm intelligence has been integrated with data mining to generate
lightweight but robust methods to detect and identify the flow of data effectively. This review paper pursues
a twofold goal. First is to review various swarm-based anomaly detection methods and to provide new
insights in that direction. Secondly, to replenish the literature with fresh reviews on swarm-based data mining
studies based on anomaly detection. Further it discusses various methods and architectures of anomaly
detection based on statistical, machine learning and data mining techniques.