Emitter Identification of Digital Modulation Transmitter Based on Nonlinearity and Modulation Distortion of Power...

Area of Science:

• Electrical Engineering and Computer Science
• Signal Processing
• Cybersecurity

Background:

• Specific Transmitter Identification (SEI) is crucial for wireless security, spectral management, and military applications, enabling non-cooperative transmitter recognition.
• Existing SEI methods often rely on data-driven approaches, limiting their ability to identify signals with previously unseen modulation parameter changes.
• Traditional methods like bispectral analysis show rapid performance degradation in low signal-to-noise ratio (SNR) environments and fail when carrier frequencies change.

Purpose of the Study:

• To develop a novel SEI technique that overcomes the limitations of existing data-oriented methods by analyzing intrinsic signal characteristics.
• To create a robust transmitter fingerprint by extracting coefficients representing modulator distortion and power amplifier nonlinearity.
• To evaluate the proposed method's performance against bispectral analysis, particularly under challenging conditions like low SNR and altered signal parameters.

Main Methods:

• Demodulation and reconstruction of digital modulation signals to isolate modulator distortion and power amplifier nonlinearity.
• Extraction of unique fingerprint characteristics (coefficients) from the reconstructed and received signals.
• Utilizing a Convolutional Neural Network (CNN) to classify these fingerprint characteristics for specific transmitter identification.

Main Results:

• The proposed method achieved 86% recognition accuracy at SNR = 0 dB, significantly outperforming bispectral methods which degraded rapidly below SNR = 20 dB.
• The technique maintained approximately 70% accuracy even when the test signal's carrier frequency was altered, a condition where bispectral features failed.
• Classification accuracy exceeded 70% for four different transmitters at SNR = 0 dB when the test signal's baud rate was changed.

Conclusions:

• The developed SEI method, leveraging CNNs to analyze signal distortion and nonlinearity, offers superior robustness compared to traditional techniques.
• This approach demonstrates significant potential for reliable transmitter identification in non-cooperative scenarios, even with low SNR and varying signal parameters.
• The findings highlight the effectiveness of using intrinsic signal distortion as unique transmitter fingerprints for enhanced wireless communication security.