Content area
Full Text
Abstract: This article hypothesizes that the presence of well-protected memory of sensor networks improves their performance as a multi-agent system. The agent's memory stores information about typical reactions to information signals from receptors, as well as information about the state of effectors and available resources. It also stores programs for processing input information into control signals supplied to effectors and the results of reactions to a particular external situation, the number and variety of knowledge and programs stored in it, the degree of development of the internal model of the external world, and reflection options that determine the complexity and nature of behavior agent. The main components of the memory block are filter systems that provide the most important information for the agent, the internal model of the outside world, and the model of the agent itself.
Purpose: to prove that using the FPGA platform guarantees programmable memory protection to maintain its level of autonomy and intelligence.
The object of study: cybersecurity of sensor networks with protected memory
Approaches: agent theory, machine learning
Objective: To prove that the use of the FPGA platform guarantees programmable memory protection to maintain its level of autonomy and intelligence.
Findings: sensor network security models
Keywords: wireless sensor network (WSN), information security, multi-agent system, machine learning, FPGA bit stream algorithms; clustering intrusion detection methods
1.Introduction
Well-known and frequently used wireless networks have long established themselves as productive technologies for many industries, agriculture, medicine and other business areas. It has become common to use WSN for manufacturing processes of various kinds. There is a lot of research and projects in these sectors, but now only one thing excites minds the minds of project developers - this is information security of networks in all aspects. So according to the data of 2018, in wireless networks are use 22 types of hacker attacks (Ahmed, 2018). The author (Chun-ta Li, 2010) argues that attackers can easily compromise the WSN by using the broadcast medium and the lack of protection against unauthorized access. This is all the more relevant in today's digital environment, cybercrime has become a disaster, for example, every 39 seconds there is a hacker attack, 300,000 new malware is created every day (Galov, 2020). In this connection, sharply there is a...