Open Access

ARTICLE

A Novel Malware Detection Method Based on IPSO-Optimized LSTM

Zheng Yang¹, Hua Zhu^1,*, Zhao Li², Gang Wang³, Meng Su¹

1 Electric Power Research Institute of Yunnan Power Grid Company Ltd., Kunming, China
2 Department of System Operation, Yunnan Power Grid Co., Ltd., Kunming, China
3 School of Electrical Engineering, Chongqing University, Chongqing, China

* Corresponding Author: Hua Zhu. Email:

Journal of Cyber Security 2026, 8, 189-210. https://doi.org/10.32604/jcs.2026.078232

Received 26 December 2025; Accepted 03 April 2026; Issue published 18 May 2026

Abstract

The rapid integration of IoT technologies in modern power systems, while enhancing operational efficiency, has introduced critical cybersecurity vulnerabilities. The proliferation of interconnected terminal devices across diverse operational domains has escalated cybersecurity risks, particularly from sophisticated malware attacks targeting critical grid infrastructure. These threats manifest through Application Programming Interface (API) call hijacking, command injection in industrial control protocols, and evasion of conventional signature-based detection systems. To address these challenges, this paper proposes a novel malware detection framework specifically designed for power IoT ecosystems. First, a malware detection model based on long short-term memory network (LSTM) is established; then, the API call sequence data is preprocessed and mined for malware feature coefficients; next, the improved particle swarm optimized (IPSO) is introduced to optimize the four hyper-parameters of the LSTM; and finally, a publicly available dataset is used in the experimental part to evaluate and validate the proposed IPSO-LSTM malware detection model is evaluated and validated by comparing it with other detection methods, and the results indicate that the proposed method shows excellent detection performance in the experiments and can realize efficient and accurate detection of malware, which provides a strong guarantee for the safe operation of power mobile terminals.

---

Keywords

---