Mathematical modeling of Internet of Things traffic in AD-HOC networks with hybrid routing

Abstract

This paper considers the impact of Internet of Things traffic, which is formed by monitoring and control systems or other systems, when the properties of this traffic are described by the properties of the regular flow. The impact of this traffic on such key QoS indicators as data delivery delay and probability of loss is estimated. As a model of communication network the system of queuing (SMO) with the combined discipline of service is considered. The analysis of trends in the development of the infocommunication system shows that the share of IP traffic in promising communication networks will significantly increase, which will lead to its impact on the quality of service .Given that the traffic in the network will also include telemetry traffic, this impact can have a significant impact on the quality of its service. Modern AD-HOC networks are based on the principle of “averaging”. According to statistics, many data streams with random variations in density will result in some average traffic. However, this approach does not work in networks prone to strong peak emissions. Such peculiar, time-localized “congestions” cause significant packet losses, even when the total demand of all flows is far from the maximum allowable values. This negatively affects the efficiency of network bandwidth utilization. The classical Poisson model of traffic, which was used in the design of network protocols, does not reflect the real reality: the data of real network traffic have the property of self-similarity.