Optimal method to monitor network for IoT devices based on anomaly detection

Abstract

Many challenges have been identified to monitor, manage, process, and store the big data that accumulates from different sources in the IoT concept. The focus of this paper is very significant and limited to solving the problem of monitoring classified big data. Detection of anomalies in a grouping of classified data made it easy to monitor and help to make decisions for action to operate. There is no need to store, process, or manage the redundant data further that is already within the range of the group. So, the main concern is abnormal values in the groups that need to be processed further and require focus. The method proposed in this paper serves as an optimal solution designed to address the visualization challenges associated with dense and high-volume datasets. Our approach involves a strategic process of categorizing data into groups and pinpointing anomalies within these groups. This systematic classification not only enhances data organization but also plays a pivotal role in simplifying the visualization of intricate data patterns. Additionally, this method brings about significant cost efficiencies by strategically optimizing the expenses incurred in processing operations and the allocation of storage space for the equipment.