?
Modeling the Processing of Non-Poissonian IIoT Traffic by Intra-Chip Routers of Network Data Processing Devices
The ecosystem of the Internet of Things (IoT) continues to grow at the present time and covers more and more areas. One of these areas is the Industrial Internet of Things (IIoT), which integrates sensors and actuators, business applications, open web applications, multimedia security systems, positioning and tracking systems. Each of these components creates its own data stream and has its own parameters of the probability distribution when transmitting information packets. One such distribution, specific to the Trumpf TruPrint 1000 IIoT system, is the beta distribution. In our article, we described the results of a study of the processing of such a data flow by an agent model of a 5 × 5 NoC switch fabric. The concepts of modern telecommunication networks (5G / 6G) imply the processing of "small" data in the place of their origin, not excluding the centralized processing of big data. This process, which involves the transmission, distribution and processing of data, involves a large number of devices: routers, multiprocessor systems, multi-core systems, etc. We assumed that the data stream is processed by a device with a network structure, such as NoC, and goes to its built-in router. We conducted a study of how the average queues of a 5 × 5 router change with changes in the parameters of a data stream that has a beta distribution.