| |
Problem statement. Modern monitoring and notification systems face a growing volume of collected metrics. The number of devices, and consequently the amount of data in such systems, has increased significantly, especially with the rise of the Internet of Things (IoT). The number of monitored objects has grown exponentially. Operators in these systems must track an increasing number of potential notifications, which increases the cognitive load on the operator. In the current context, a monitoring system must not only process large volumes of data quickly but also present it in an intuitive and comprehensible form to reduce the user’s cognitive burden.
Purpose. The primary goal of this research is to analyze the processes within monitoring and notification systems, including information flows, load, performance, timing, and frequency of processes, with attention to the cognitive load on human operators and the characteristics of the information delivered to them.
Results. Information flows in monitoring systems and their subprocesses have been analyzed. A test environment was built to simulate the increasing number of monitored objects and notifications. The experiments demonstrated that sending notifications exclusively upon state changes reduces redundancy in notifications and aligns them with the information entropy curve, representing the amount of information. Adjusting notification conditions toward less equiprobable triggers significantly reduces the number of necessary notifications. Introducing a waiting period for condition checks reduces the operator's cognitive load but increases the system's reaction time. As the number of monitored objects grows, optimizing the data collection frequency minimizes overall system delays in detecting and sending notifications while reducing the operator's workload associated with processing and extrapolating outdated alerts.
Practical significance. The described criteria and methods for ensuring effective settings have practical significance for configuring the frequency of monitoring system subprocesses, designing notification conditions, and implementing mechanisms for their logging and delivery. These findings are applicable to monitoring systems in industrial and IT infrastructure contexts.
Keywords:metric, monitoring, notification, monitoring system, notification system, cognitive load, information entropy, monitored object
|
LEGAL
INFORMATION:
