Relay Selection Machine Learning-Based for a DF Cooperative System with Energy Harvesting and Signal Space Diversity

Machine learning techniques have been employed in wireless communication systems to offer resilient and low-complexity solutions. Accordingly, this paper explores the utilization of machine learning algorithms for real-time relay selection in a cooperative communication system with signal space diversity and multiple energy harvesting relays. The assumed decoding scheme at the relays is decode-and-forward (DF), and the relay selection criteria involve successful decoding from the source, sufficient energy, and the best channel to the destination. Con-ventional machine learning algorithms, including Decision Tree (DR), Logistic Regression (LR), K-Nearest Neighbors (KNN), and Support Vector Machines (SVM), are implemented, and the algorithm with the highest accuracy is utilized for real time relay selection. Performance evaluation is in terms of outage probability, and the exact outage probability performance is de-termined through Monte Carlo simulation. Results indicate that KNN exhibits the highest accuracy; nevertheless, its performance falls short of closely approximating the exact outage probability due to the complexity of the data, necessitating the need for more advanced classification models such as artificial neural networks or deep neural networks to accurately approximate the exact performance of the proposed system.