Automatic Modulation Classification for MASK, MPSK, and MQAM Signals Based on Hierarchical Self-Organizing Map

Area of Science:

• Electrical Engineering
• Signal Processing
• Machine Learning

Background:

• Automatic modulation classification (AMC) is crucial for modern communication systems.
• Current clustering models struggle with a high number of modulation types, accuracy, and computational load.
• Distinguishing between various amplitude-shift keying (MASK), phase-shift keying (MPSK), and quadrature amplitude keying (MQAM) signals presents a challenge.

Purpose of the Study:

• To develop a more accurate and computationally efficient AMC method.
• To improve the classification of multiple MASK, MPSK, and MQAM signals.
• To overcome the limitations of existing clustering models in handling complex modulation schemes.

Main Methods:

• A hierarchical self-organizing map (SOM) was proposed, utilizing high-order cumulants (HOC) and amplitude moment features.
• A two-layered SOM architecture was employed for progressive clustering from rough to fine distinctions.
• A discrete transformation with modified activation functions was implemented for enhanced MQAM signal clustering.

Main Results:

• The hierarchical SOM effectively clustered various MASK, MPSK, and MQAM signals without requiring predefined thresholds.
• The discrete transformation improved cluster boundaries and classification accuracy for MQAM signals.
• The proposed model demonstrated superior performance compared to existing clustering methods in terms of accuracy and resource utilization.

Conclusions:

• The hierarchical SOM offers a robust and efficient solution for automatic modulation classification.
• This method successfully handles a larger number of modulation categories with higher accuracy.
• The approach reduces computational resources, making it suitable for practical communication systems.