Determining the average service time of the package on the Internet of things

Abstract

The Internet of Things as a complex system with delays in signal and control information can predict its state and solve problems of service quality management in real time. The described interdependence of network parameters of different levels allows us to consider the impact on the network. In general, we can say that changes in this network occur due to certain factors that affect the values of network parameters, and these actions can affect any network at any level, which usually leads to changes throughout the network. A representation for estimating the average value of the packet service time in the Internet of Things is proposed, a model of station behavior in the form of a Markov chain with a discrete integer time, the unit of which is a virtual slot, is constructed. Based on the accepted model, we should consider 3 types of virtual slots: «empty», «successful» and «collision» slot. For these cases, we obtained the basic probability calculation formulas. Demographic parameters, territorial parameters and a generalized indicator, which includes many events affecting the network, were chosen as the main external influences. Changes that lead to the influence of external factors on the subsystem may be some actions to modernize the network or vice versa - failure of any element of the subsystem. However, when running a distributed application, there is a conflict between the balanced distribution of objects on the processors and the low speed of messaging between processors. If logical processes are distributed between processors in such a way that the communication costs between them are reduced to zero, then some processors may be idle, while others will be overloaded. Otherwise, a «well-balanced» system will require high communication costs. Therefore, the balancing strategy should be such that the computing nodes are loaded fairly evenly, but the communication environment should not be overloaded. This description of the functioning of the Internet of Things allows you to predict the state of the network to choose appropriate management solutions to improve the forecast situation.