AI-Based Edge Node Protection for Optimizing Security in Edge Computing
Applications of AI for Interdisciplinary Research, CRC Press, ss.218-230, 2024
- Yayın Türü: Kitapta Bölüm / Araştırma Kitabı
- Basım Tarihi: 2024
- Doi Numarası: 10.1201/9781003467199-19
- Yayınevi: CRC Press
- Sayfa Sayıları: ss.218-230
- Boğaziçi Üniversitesi Adresli: Evet
Özet
Edge computing has brought improvements that reduce latency and unnecessary bandwidth usage by bringing processing power closer to the source where data are produced. However, due to the decentralized nature of edge computing, edge nodes are vulnerable to attacks. For this reason, some innovative solutions are needed to ensure the security of edge nodes. One of these solutions can be artificial intelligence (AI)-based biometric recognition applications, which have recently focused on by researchers. In these applications, biometric data such as fingerprints and facial recognition obtained from users are used to ensure that only authorized users can access edge devices. On the other hand, the fact that biometric data carries unique information about the person means that it will cause privacy problems if stolen. In this chapter, we present AIBioSec, a new framework that uses face recognition in edge nodes to solve this problem and is also based on cryptographic encryption to secure biometric data. In AIBioSec, AES, DES, and 3DES cryptographic algorithms were compared according to encryption and decryption times. The results showed that AES-128 works faster than other algorithms in securing biometric data, with 1007 ms for encryption and 1137 ms for decryption.