STATISTICAL ANALYSIS OF REAL TRAFFIC OF MACHINE-TO-MACHINE COMMUNICATION (M2M)

Abstract

The development of digital technology has spawned the concept of the Internet of Things (IoT). The concept basis is the machine-to-machine interaction (M2M) technology, which allows devices to exchange information. The most effective data transmission medium for M2M devices is mobile communications. Rapid growth of machine-to-machine М2М traffic in mobile communication network defines the actuality of the research problem, its features and characteristics. Research outcomes are indispensable at the network modeling, planning, analyzing the М2М traffic impact at quality of service (QoS) of mobile network communication. The article analyzes the real traffic in the LoraWan network. Aggregated traffic coming to the network server from all devices is considered. To model the М2М batch traffic, apart from specifying the statistic characteristics it is necessary to assess its self-similarity. In order to define the traffic self-similarity there has been computed Hurst parameter. On the basis of STATISTICA programs batch we have conducted statistical analysis and short-term forecasting of real М2М traffic by method of exponential smoothing.