Thesis is devoted to solving the urgent and important scientific and technical task of developing stochastic information technology for monitoring environmental parameters in the modern Internet of Thing (IoT) concept, taking into account the a priori uncertainty of the sources of information and the possibility of emerging crisis situations.
The analysis of principles of construction, technological decisions and directions of development of monitoring systems in the IoT concept was carried out. As a result of this analysis shortcomings of known approaches were identified and the necessity of creating mathematical models, methods, communication protocols of WSN networks with random access and corresponding information monitoring technologies for ensuring high productivity, quality and the vitality of their functioning was proved.
It has been established that the IoT concept has three interrelated basic issues: providing information security (IoT Security), scaling up the growing volume of technical devices and data (IoT Scalability), and also IoT Technical Solutions and Low-Power Consumption. Also, the analysis of protocols for solving IoT tasks was carried out: 1) MQTT: protocol for collecting data of devices and transmitting their servers (D2S); 2) XMPP: protocol for connecting devices to humans, partial case of D2S-schemes when people connect to servers; 3) DDS: fast bus for integrating smart devices (D2D); 4) AMQP: The system organizes queues for connecting servers to each other (S2S).
Stochastic models of the functioning of wireless sensor networks that use randomized network parameters (with variable number of nodes and random participation of nodes in separate groups of network nodes) have been improved. It allowed to estimate the probability of collision of signals and to more effectively design communications protocols of the IoT. These models allowed to estimate the probability of collision of signals: the maximum number of nodes that provide the quality of transmission at the level of the probability of collision no higher than 10-2 is 50, with the number of nodes involved in the collision is negligible in comparison with the average number of transmissions, in particular the ratio of the average number involved in the collision of nodes to the average number of transmissions is 10-7.
The method of monitoring environmental parameters has been improved. It takes into account the unsteady spatial and temporal localization of primary sources of measurement and optimization of the dynamic monitoring process, which made it possible to ensure the timely and prompt receipt of information from primary sources of information with specified quality indicators for effective management decision-making.
The development of information monitoring technology, which, due to the use of stochastic models of the operation of wireless sensor networks and the advanced monitoring method, has allowed the software-technical complex (using Arduino, JavaScript, NodeJs, HTML and CSS) to monitor real-time environment parameters in the modern IoT concept. The designated software-technical complex for monitoring of real-time environmental parameters can be used as a prototype for monitoring organization in dynamically changing environments and in case of emergencies of a different nature (in various spheres).
The experimental research of proposed models, method and information technology of monitoring was carried out. It has been found that when transmitting data at short distances (for example, indoors – laboratory, office, home), devices can use the PAN provided by wireless technologies such as BLE (Bluetooth Low Energy), ZigBee, 6LoWPAN and the leading USB interface. When it comes to transmitting data over long distances (for example, in an large office or in a large building), you can use a Local Area Network. Wired LAN in most cases is based on Ethernet and fiber optic technology, and wireless one can be based on Wi-Fi technology. Also can be used WiMAX, LTE, and LPWAN to organize global WAN. Measured parameters are not limited to humidity and air temperature (as it was performed in experimental part of this thesis). Any parameters can be used in view of available sensors, which are part of the software-technical complex.
Besides, specialized UML diagrams of simulation precedents and simulation sequences of the proposed information technology have been developed. The results of the dissertation have been used and implemented at the National Aviation University, the Central Ukrainian National Technical University and the telecommunication company Local Students Networks (special implementation acts are attached to thesis).
