Adaptive scheduling of high bandwidth periodic packet switch sources for congestion handling

Abstract

Industrial Internet of Things (IIoT) networks are quite prevalent in many types of sectors today. Also, in the near future, many sensors are estimated to be connected and used. With the development of sensor technologies, they transmit extra data to the core network. Optimization methods are crucial to cater to this high bandwidth traffic without expanding the aggregation servers and links. This work has proposed a two-tier IIoT network architecture with several scheduling algorithms at tier one (Data Loggers). The network has been simulated using Omnet++. The traffic distribution at the aggregation layer has been compared for several scheduling algorithms to get the best possible method to optimize the network without using a central- ized scheduler or controller. The results show that out of the proposed algorithms, the best methods in terms of distributing traffic are Initial Uniform Slot Selection (IUSS) and Initial Bernoulli Slot Selection (IBSS) methods.