Accelerated neural intrusion detection for wireless sensor networks

Abstract

Wireless sensor networks (WSNs) form an important layer of technology used in smart cities, intelligent transportation systems, Industry, Energy, Agriculture 4.0, the Internet of Things, and, for example, fog and edge computing. Cybernetic security of such systems is a major issue and efficient methods to improve their security and reliability are sought. Intrusion detection systems (IDSs) automatically detect malicious network traffic, classify cybernetic attacks, and protect systems and their users. Neural networks are used by a variety of intrusion detection systems. Their efficient use in WSNs requires both learning and optimization and very efficient implementation of the detection. In this work, the acceleration of a neural intrusion detection model, developed specifically for wireless sensor networks, is proposed, studied, and evaluated.