Multi-objective optimization algorithms for intrusion detection systems in IoT: a systematic review and future research directions

Abstract

With the increasing use of the Internet, massive amounts of data are transferred between various communication devices. Network security is a critical area of research in today’s network environment, as data must be transmitted securely between connected devices. The digital revolution and the expansion of operations have amplified the importance of intrusion detection systems (IDSs) in networks. Numerous research concepts have been proposed that utilize swarm algorithms, evolutionary algorithms, deep learning, and machine learning techniques in the field of IDSs. IDSs must be chosen among several factors, including false alarm rate, relevance, detection accuracy, and other goals. To identify attempts to exploit security vulnerabilities and mitigate the likelihood of network security threats, this study offers a thorough systematic analysis of intrusion detection (ID) in Internet of Things (IoT) systems leveraging multi-objective optimization algorithms (MOAs). For extremely complicated IoT networks, MOAs offer a range of efficient options for IDSs. This study identifies various ID objectives and conducts a comparative examination of multi-objective intrusion detection approaches in the IoT, based on their strategies and datasets utilized in the assessment criteria. This survey is a representative analysis conducted between 2018 and 2025 in the field of IDSs. The promising potential in IoT networks to improve several competing ID objectives is illustrated by MOAs. This study also discusses current challenges and future avenues for further research. It aims to shed light on research gaps that can be filled when developing IDSs for IoT networks, as well as to showcase the latest advancements in ID approaches. The research community and novice cybersecurity researchers can use this study as a benchmark to understand and develop effective IDS models.

The authors' abstract, as published at the source. Artificial Intelligence Review, 2026 · DOI ↗