APPLICATION OF FUZZY MODELS TO CONTROL THE STREAM OF MESSAGES IN DATA TRANSMISSION INFRASTRUCTURES

Abstract

The report presents one of the problems that arises in distributed data transmission infrastructures - node overload in data transmission when there is unstable communication between nodes. As a solution, a mechanism for controlling the message flow intensity based on a fuzzy model is proposed. The developed fuzzy control model has demonstrated its effectiveness in the network for collecting and transmitting quotes of financial instruments from various trading platforms to the information-analytical system of a financial company. Message queues at the sources have significantly decreased, while minimizing losses of quotes on unstable communication channels. Additionally, the possibility of more flexible adjustment of fuzzy control by expanding the number of metrics of input parameters of the fuzzy model has been explored.