What is it about?
Indoor navigation systems rely on Wi-Fi signals to determine locations, but hackers can manipulate these signals to create false locations, misdirect people, or disrupt critical services in hospitals, airports, and security-sensitive areas. These adversarial attacks can pose serious risks, making indoor localization systems unreliable and unsafe for real-world use. SENTINEL introduces a new AI-driven defense using capsule neural networks, which recognize and resist these attacks by learning deeper relationships in signal patterns. Unlike traditional models that can be easily tricked, SENTINEL’s approach ensures up to 3.5x better accuracy under attack, making indoor positioning systems more secure, trustworthy, and resilient against manipulation.
Featured Image
Photo by Brett Jordan on Unsplash
Why is it important?
Hackers can manipulate indoor navigation systems in hospitals, airports, and security-sensitive areas, creating false locations and disrupting operations. This poses serious risks for emergency response, patient tracking, and smart infrastructure. SENTINEL defends against these attacks using capsule neural networks, which detect and resist manipulation, ensuring accurate and secure indoor localization even in high-risk environments.
Perspectives
As indoor localization becomes critical for healthcare, security, and smart infrastructure, ensuring resilience against cyber threats is more important than ever. SENTINEL’s ability to detect and resist adversarial attacks makes it a key step toward secure and trustworthy indoor positioning systems. By leveraging capsule neural networks, this research ensures long-term reliability in environments where accuracy and security are crucial, paving the way for safer, more robust navigation technologies.
Danish Gufran
Colorado State University
Read the Original
This page is a summary of: SENTINEL: Securing Indoor Localization Against Adversarial Attacks With Capsule Neural Networks, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, November 2024, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/tcad.2024.3446717.
You can read the full text:
Contributors
The following have contributed to this page
