Comprehensive Analysis of IoT Malware Evasion Techniques

A Al-Marghilani

doi:10.48084/etasr.4296

Abstract

Malware detection in Internet of Things (IoT) devices is a great challenge, as these devices lack certain characteristics such as homogeneity and security. Malware is malicious software that affects a system as it can steal sensitive information, slow its speed, cause frequent hangs, and disrupt operations. The most common malware types are adware, computer viruses, spyware, trojans, worms, rootkits, key loggers, botnets, and ransomware. Malware detection is critical for a system's security. Many security researchers have studied the IoT malware detection domain. Many studies proposed the static or dynamic analysis on IoT malware detection. This paper presents a survey of IoT malware evasion techniques, reviewing and discussing various researches. Malware uses a few common evasion techniques such as user interaction, environmental awareness, stegosploit, domain and IP identification, code obfuscation, code encryption, timing, and code compression. A comparative analysis was conducted pointing various advantages and disadvantages. This study provides guidelines on IoT malware evasion techniques.

Highlights

Internet of Things (IoT) is a system where various interconnected objects transfer data over a wireless network without requiring any human intervention [1, 2]
This study aims to analyze the malware detection techniques in IoT applications, compare the existing studies on IoT malware evasion techniques highlighting their advantages and limitations, and identify the various malware evasion techniques
Security researchers use machine learning techniques to deal with these security issues and enhance IoT devices' efficiency

Summary

INTRODUCTION

Internet of Things (IoT) is a system where various interconnected objects transfer data over a wireless network without requiring any human intervention [1, 2]. Technologies, and software are embedded into these devices to connect and transfer data within the network [3]. As IoT continues to grow, a critical issue is emerging regarding the privacy and the security of data transferred over the network. In behavior-based malware detection, the object is evaluated based on its actions [26] before executing actions to analyze suspicious activities and get rid of threats. This technique is further classified as static, dynamic, and hybrid. Signature-based malware detection is working with known threats and detects them by establishing unique identifiers It is classified as static, dynamic, and hybrid [28]. This study aims to analyze the malware detection techniques in IoT applications, compare the existing studies on IoT malware evasion techniques highlighting their advantages and limitations, and identify the various malware evasion techniques

SECURITY ISSUES IN IOT APPLICATIONS

CHALLENGES AND ASSOCIATED TECHNIQUES TO

VARIOUS MALWARE ATTACKS IN IOT

VARIOUS MALWARE EVASION TECHNIQUES IN IOT

CONCLUSION