The adoption of Internet of Things (IoT) practices holds great potential for the industry. Nonetheless, ensuring reliability and maintaining high-quality service is paramount

Monitoring Approach. Two distinct techniques are available for network monitoring: active and passive methods [38]. The chosen approach for the proposed monitoring system involves active measurements. It employs active probing by transmitting internet control message protocol (ICMP) packets across the network to remote endpoints. ICMP packets are often filtered by routers due to security concerns. However, since all remote endpoints are registered on a private APN, ICMP packets can still be used to assess the stability of the connection. The monitoring system can be configured with one of three restriction profiles: server, administrator, or viewer. The server profile grants the system full access to execute active probing procedures, with the condition that the system must have access to the APN. The administrator profile permits the addition or removal of remote endpoints but does not authorize the execution of active probing procedures. The viewer restriction profile only allows for viewing the current network status, preventing the execution of active probing methods while in viewing mode.
The monitoring system has been designed with modularity in mind, making it straightforward to add additional endpoints for monitoring purposes. To optimize the monitoring system's performance, multithreading techniques have been employed for all software routines. This ensures effective monitoring of remote endpoints, regardless of their number. Figure 4 illustrates the primary software flow of the monitoring system. The software routine is executed periodically in real-time. Since data is transmitted from each site daily, the execution frequency can be narrowed down to hourly intervals, preventing network flooding with ICMP packets. This interval is adjustable as needed. Real-time monitoring requires an active connection to the relational database. Therefore, a database connection check is the first step. Without a successful database connection, the monitoring routine is not executed, and no data is displayed to the user. If a database connection is established, the host machine's status determines whether the monitoring routine proceeds. When the host machine is idle, all remote IP addresses are retrieved from the database and stored in a data structure. Worker threads are then created, with each responsible for checking the connection status of a single IP address. Through active probing, all active remote endpoints will respond, and the resulting replies are stored in the database for processing and display purposes.