Introduction

A malware probe attack is a cyber-attack that involves an attacker attempting to gain access
to a system or network. A probe is the initial stage in a malicious attack, where the attacker
attempts to map out the capabilities and vulnerabilities of a system or network. This can be
done by sending special requests (probes) to the target system. (S. Mendhurwar and R.
Mishra 2021) describe the expansion of big data and the evolution of IoT in a smart city.
(Francesca et al. 2019) surveyed the security risks in IoT systems and discussed
counteractions. Malware probes are often used by hackers to gain access to systems, as they
allow them to try out different techniques and see how they work on both internal and
external systems. (Luqman et al. 2021) performed an SLR based on the privacy of the IoT
system. It identified challenges with regard to the privacy of the IoT system exposed, type of
attacks occur in the IoT system and recommendations for future studies. The Malware probe
prediction is used to exploit the vulnerabilities of IoT devices to penetrate into their
environment and perform malicious activities.

The research work intends to analyse and categorize the network traffic of IoT devices for the
detection of sophisticated ever-growing cyber-attacks with high detection accuracy in a low
detection time. Previously the research papers have focused on malware probes and their
various uses in IoT devices but have less efficiency. Introduction of an intrusion detection
system using Support Vector Domain to determine if network traffic is an attack or not, a
novelty detection model based on SSPV-SVDD as a classifier and SMO as a solution is used
(Costa et al. 2018). An approach to detect ransomware using power consumption K-Nearest
Neighbors, Neural Networks, Support Vector Machine, and Random Forest have been used
in this method for classification (Azmoodeh et al. 2018). Two layers dimensionality reduction
method and Two-tier Classification (TDTC) classification module for detection of the
malicious activity, i.e., U2R User to Root and Remote to Local R2L (Pajouh et al. 2016).
Potential users would then be able to mitigate from attacks such as video injection etc. LSTM
is used to hunt IoT malware-based on Opcodes. The evaluation of the proposed model is
conducted using ARM-based IoT application execution Opcodes. The feature selection
technique text mining is used in this method to obtain an important feature vector from
Opcode (HaddadPajouhet al. 2018). The goal of the attacker is to exploit the vulnerabilities of
IoT devices to penetrate into their environment and perform malicious activities.