A MODEL OF SITUATIONAL CONTROL OF THE TELECOMMUNICATION AERIAL PLATFORM FLIGHT TRAJECTORY TO COLLECT DATA FROM NODES OF A WIRELESS SENSOR NETWORK

Abstract

A telecommunication aerial platform (TA) based on an unmanned aerial vehicle is considered a promising technology for data collection in wireless sensor networks in the absence of connectivity between network nodes and any communication infrastructure. In fact, TA acts as a mobile gateway and could collect data from several nodes at the same time. The trajectory of its movement, location of points and data exchange intervals significantly affect the efficiency of the data collection process. The article considers a model of situational control of flight trajectory construction for data collection to achieve certain target functions: optimization of data collection time and network operation time. In the work, a set of situations on the network is classified and corresponding production rules for building a TA trajectory are defined, which implement a defined decision-making hierarchy: network, cluster, TA, node, considering the target management functions. At the network level, the rules for determining the number and size of clusters are applied, and a basic solution for determining the collection points and the trajectory of their flight is being built. At the level of each cluster, the TA adjusts the basic solution during the flight based on considering the parameters of the actual state of the cluster nodes. At the level of TA-node interaction, the energy consumption of the node and the speed of data transmission are optimized by reducing the distance of the node to the telecommunication aerial platform. To reduce the number of rules, meta-rules are proposed. This approach allows you to achieve optimization of the target functions of the data collection process and ensure decision-making in real time. The simulation results proved the possibility of reducing data collection time by 10–15 % or increasing network operation time by 12–17 % compared to existing solutions.