The application of intelligent data analysis methods for traffic processing in a wireless sensor network

Abstract

Wireless sensor networks (WSNs) enable the collection and transmission of data in the Internet of Things (IoT), and are widely used in various sectors, such as the military, industrial, agricultural, environmental, transport, social and medical sectors. WSNs are networks that provide continuous communication and access to information at any time and in any place. The main areas of research in this work relate to the application of intelligent data analysis methods to improve data transmission characteristics and quality of service in a heterogeneous wireless sensor network, which consists of a ground segment with stationary sensor nodes and an aerial network; this is an extremely pressing challenge in the monitoring and control of coverage areas using the Internet of Things. This paper investigates a method for efficient clustering in a ground-based wireless sensor network, which ensures the identification and allocation of an optimal number of clusters and the designation of primary cluster nodes, thereby ensuring efficient processing and transmission of traffic across different network segments. Approaches have been proposed and are continuing to be developed that combine various methods of cluster formation and data transmission between nodes of different topological types. Data can be sent from an end node directly to the base station, via a primary cluster node, or via a sequence of such nodes before reaching the base station. In solving the tasks set, methods from graph theory, artificial neural networks, genetic algorithms and K-means, as well as software engineering, were utilized.