Radio Frequency Fingerprint-Based Intelligent Mobile Edge Computing for Internet of Things Authentication

Area of Science:

• Computer Science
• Electrical Engineering
• Cybersecurity

Background:

• Resource-constrained terminals in mobile edge computing (MEC) require robust authentication methods.
• Traditional encryption-based security can be computationally expensive for edge devices.
• Radio Frequency Fingerprinting Identification (RFFID) offers a promising alternative for device authentication.

Purpose of the Study:

• To propose a lightweight RFFID scheme for authenticating numerous resource-constrained terminals in MEC environments.
• To develop a two-layer model that leverages both edge and cloud computing for efficient authentication.
• To enhance authentication rates without relying on encryption-based techniques.

Main Methods:

• A two-layer architecture: MEC devices handle initial signal collection, feature extraction, and access decisions.
• Remote cloud performs advanced machine learning for feature learning, model generation, and recognition.
• Effective utilization of wavelet features for improved RFFID accuracy.

Main Results:

• The proposed RFFID scheme achieves higher recognition rates compared to traditional methods.
• The two-layer model effectively utilizes MEC and cloud resources for authentication.
• Demonstrated efficiency in Internet of Things (IoT) application scenarios.

Conclusions:

• The novel lightweight RFFID scheme is efficient and effective for authenticating devices in MEC.
• The two-layer approach combined with machine learning and wavelet features significantly improves authentication performance.
• This method provides a viable, non-encryption-based solution for securing resource-constrained terminals.