Energy-efficient transmit antenna selection with Fast-ABC-Boost.

Area of Science:

• Wireless communication
• Signal processing
• Machine learning

Background:

• Multiple-Input Multiple-Output (MIMO) systems offer enhanced data rates but face energy consumption challenges.
• Antenna selection is a key strategy for improving energy efficiency (EE) in MIMO systems.
• Existing methods for antenna selection often involve complex optimization or learning algorithms.

Purpose of the Study:

• To develop a novel, energy-efficient antenna selection technique for MIMO systems.
• To address the antenna selection problem as a multi-class classification task.
• To maximize energy efficiency (EE) in MIMO systems through advanced classification.

Main Methods:

• An antenna selection technique based on Fast-Adaptive Base Class-Boost (Fast-ABC-Boost) was proposed.
• Fast-ABC-Boost enhances the performance of weak classifiers, such as regression trees, to form a robust decision committee.
• The technique frames antenna selection as a multi-class classification problem for EE maximization.

Main Results:

• Fast-ABC-Boost demonstrated superior energy efficiency (EE) performance compared to existing methods.
• The proposed technique showed better EE performance than Deep Reinforcement Learning (DRL) and Cyclic Binary Particle Swarm Optimization (CBPSO).
• The method achieves high EE performance with feasible computational complexity.

Conclusions:

• Fast-ABC-Boost offers a highly effective and energy-efficient solution for antenna selection in MIMO systems.
• The classification-based approach provides a competitive alternative to conventional optimization and DRL methods.
• This technique presents a promising direction for developing energy-efficient wireless communication systems.