A Survey on A Hybrid CNN + Transformer with Genetic Algorithm for Intrusion Detection in Wireless IoT Devices
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Wireless IoT devices are all around us in smart homes, hospitals, factories, and fields quietly sensing and sharing data to make life easier and safer.But because they run on tiny batteries, have almost no memory, and talk over open airwaves, they are incredibly easy to attack.Traditional security tools either kill the battery or fail against new threats no one has seen before.This is why researchers have been building smarter, self-learning intrusion detection systems that can actually live on these constrained devices.Today's best designs mix Convolutional Neural Networks (which catch suspicious patterns inside single packets) with Transformers (which understand how traffic behaves over time) and use attention to focus only on the few events that matter.To keep everything small and fast, Genetic Algorithms automatically try thousands of combinations and pick the leanest, most effective version.Even with years of progress, truly practical, battery-friendly systems for real wireless IoT networks have stayed rare.This survey traces that journey and presents our complete, ready-to-use solution: a compact CNN-Transformer with attention, optimized by genetic algorithm, and wrapped in a live Streamlit dashboard that lets anyone monitor and protect their network in real time.
Fields of Study & Subjects
Network Security and Intrusion Detection
Network Packet Processing and Optimization
Internet Traffic Analysis and Secure E-voting
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 223–237
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter