Hybrid radar emitter recognition based on rough k-means classifier and relevance vector machine

Area of Science:

• Electrical Engineering
• Signal Processing
• Machine Learning

Background:

• Increasing complexity of electromagnetic signals poses challenges for radar emitter signal recognition.
• Traditional methods struggle with accurate and efficient classification of diverse radar signals.

Purpose of the Study:

• To develop a hybrid recognition approach for improved radar emitter signal classification.
• To enhance accuracy and reduce computational complexity in radar signal identification.

Main Methods:

• A two-step hybrid approach: primary signal recognition and advanced signal recognition.
• Utilizing a novel rough k-means classifier to cluster radar emitter signal samples into certain, rough, and uncertain regions.
• Employing a relevance vector machine (RVM) trained on samples from the rough region to classify samples in the uncertain region.

Main Results:

• The hybrid approach demonstrates higher classification accuracy compared to traditional methods.
• The proposed method exhibits lower computational complexity.
• Effective classification of radar emitter signals is achieved by leveraging sample separability.

Conclusions:

• The hybrid recognition approach offers a more accurate and computationally efficient solution for radar emitter signal identification.
• The combination of rough k-means and RVM effectively handles complex electromagnetic signal environments.
• This method advances the field of radar signal processing and electronic warfare.