Secure and Intelligent Single-Channel Blind Source Separation via Adaptive Variational Mode Decomposition with...

Area of Science:

• Intelligent Systems
• Signal Processing
• Wireless Communication

Background:

• Emerging intelligent systems require robust signal processing for reliable operation.
• Single-channel blind source separation (SCBSS) is crucial for mixed and corrupted signals in wireless communication and sensor networks.
• Variational Mode Decomposition (VMD) is effective for SCBSS but sensitive to parameter selection (k and α).

Purpose of the Study:

• To develop a secure and intelligent SCBSS algorithm.
• To optimize VMD parameters (k and α) adaptively for improved performance.
• To enhance source isolation fidelity in challenging signal environments.

Main Methods:

• Proposed a secure and intelligent SCBSS algorithm using adaptive VMD.
• Optimized VMD parameters (k and α) with Improved Particle Swarm Optimization (IPSO).
• Applied improved Fast Independent Component Analysis (IFastICA) for source separation.

Main Results:

• Achieved a 15.7% improvement in separation efficiency compared to conventional methods.
• Demonstrated robust performance for BPSK and QPSK signals.
• Obtained correlation coefficients above 0.9 and signal-to-noise ratio (SNR) improvements up to 24.66 dB.

Conclusions:

• The proposed IPSO-optimized VMD and IFastICA method offers a secure and effective solution for SCBSS.
• Adaptive parameter optimization significantly enhances VMD's noise filtering and signal separation capabilities.
• The algorithm provides high-fidelity source isolation, crucial for intelligent systems operating in noisy conditions.