In the ever-evolving landscape of cybersecurity, researchers from Shanghai Maritime University have developed a novel intrusion detection system tailored for the Internet of Things (IoT). This system, detailed in a recent study published in ‘Jisuanji gongcheng’ (translated as ‘Computer Engineering’), addresses a critical gap in conventional intrusion detection methods: the inability to detect new types of attacks in real-time.
The lead author, LIU Qiang, along with colleagues ZHANG Ying, ZHOU Weixiang, JIANG Xiantao, ZHOU Weina, and ZHOU Mouguo, proposed a method that combines Stacked Sparse Autoencoders (SSAE) and Self-Organizing Incremental Neural Networks (SOINN). The SSAE performs sparse coding and feature extraction on known attack samples, while the SOINN forms a topological structure that mirrors the spatial distribution of traffic characteristics. When new attack types emerge, the SOINN adaptively generates new nodes to establish a new local topology, ensuring the system can learn and adapt without forgetting old data.
“This method can perform incremental training on the model based on new categories of data without requiring historical data,” said LIU Qiang, highlighting the system’s ability to learn continuously. The researchers demonstrated the system’s effectiveness using the CIC-IDS2017 dataset, achieving an initial accuracy rate of 98.15% and an overall accuracy rate of 57.34% after three stages of category incremental learning.
For the maritime sector, the implications are significant. IoT devices are increasingly integral to shipping operations, from smart containers to autonomous vessels. However, these devices also present attractive targets for cybercriminals. A robust intrusion detection system that can adapt to new threats in real-time is crucial for safeguarding these assets.
The commercial impact of this research is twofold. Firstly, it offers a solution for maritime companies looking to secure their IoT infrastructure. Secondly, it presents an opportunity for tech firms to develop and market adaptive intrusion detection systems tailored to the maritime industry.
As the maritime sector continues to embrace digital transformation, the need for advanced cybersecurity measures becomes ever more pressing. This research from Shanghai Maritime University represents a significant step forward in addressing this challenge, offering a glimpse into the future of maritime cybersecurity.