A Survey on Deep Learning Models for Cyber Threat Detection Systems

EasyChair Preprint 9849

Abstract

New organizational innovations and ideal model implementations are making present interruption detection and protection approaches outdated. These improvements will require new methodologies. Future technology will create huge volumes of data at lightning-fast speeds. Network safety frameworks will need to adapt to meet the expanding number of new needs. Conventional Internet-based techniques of managing data and information are being phased out in favor of electronic device applications. As a result, the information and data at issue are exposed to attacks meant to steal or destroy them. Each attack might bring down the entire system. This paper presents a deep-learning-based cyber-attack detection method for wireless sensor networks (WSN). This approach considers WSN node operations and MQTT data transport capabilities. This solution uses a deep learning model overview rather than a regular machine learning model, which improves detection accuracy. Deep learning models that use network stream data may identify network digital risks, which are cyberthreats.

Keyphrases: Artificial Intelligence, Botnets, Forensics, Intrusion, Particle Swarm Optimisation, deep learning, machine learning, neural network

@booklet{EasyChair:9849,
  author    = {Tolulope Olufemi and Wilson Sakpere},
  title     = {A Survey on Deep Learning Models for Cyber Threat Detection Systems},
  howpublished = {EasyChair Preprint 9849},
  year      = {EasyChair, 2023}}