Modeling of IIoT Traffic Processing by Intra-Chip NoC Routers of 5G/6G Networks

The concept of 5G/6G networks is widely used to design networks of Internet of Things (IoT) and Industrial IoT (IIoT) devices. The characteristics of IIoT traffic can differ significantly from the characteristics of voice traffic, so models are required to analyze IIoT traffic. In this paper, we examined the features of IIoT traffic: the Gaussian model of WSN networks, ON/OFF, which is used for target tracking sensors, models with heavy tails (Pareto, Weibull, lognormal distribution, etc.), self-similar (fractal) models - based on normalized Fractal Brownian Motion (fBM) and Fractal Gaussian Noise (fGN), as well as the IIoT traffic model. The models are obtained empirically and approximated by known models of probability distributions. We investigated an exponential distribution IIoT traffic model. In addition, we considered the prospects of using switch fabric of NoC technology and came to the conclusion that decentralization of the process of establishing connections in switch fabric has significant advantages: it allows you to flexibly distribute incoming traffic; perform distributed arbitration; flexibly scale the communication system, with a large traffic flow from one of the system inputs, to split it by assigning many different routes for its transmission; improves system reliability by allowing reconfiguration of the switch fabric. We model the processing of traffic from such systems as OBS, 1C web systems, OWM, OSM by parallel switch fabric of NoC technology. We used the previously created simulation model of the NoC 5×5 switching system and simulated the processing of traffic having an exponential distribution with different parameter values and different packet processing times.