Simultaneous Estimation of Azimuth and Elevation Angles Using a Decision Tree-Based Method.

Area of Science:

• Signal Processing
• Machine Learning
• Antenna Theory

Background:

• Accurate estimation of signal angles is crucial for antenna array systems.
• Existing methods like MUltiple SIgnal Classification (MUSIC) have limitations in speed and accuracy.
• Machine Learning offers potential for improved Direction of Arrival (DOA) prediction.

Purpose of the Study:

• To develop and evaluate a Machine Learning-based approach for predicting azimuth and elevation angles of incoming signals.
• To assess the generalization capability of the proposed model in diverse conditions.
• To compare the performance of the proposed model against the state-of-the-art MUSIC algorithm.

Main Methods:

• A Decision Tree (DT) model was trained using signal information received by an antenna array.
• The DT model was designed to simultaneously estimate azimuth and elevation angles.
• Simulations were conducted to test the model's robustness and generalization.

Main Results:

• The DT-based method demonstrated robust DOA estimation, generalizing well to unseen conditions.
• The proposed DT model achieved over 90% reduction in prediction error compared to MUSIC.
• Prediction time was reduced by approximately 50% using the DT-based approach.

Conclusions:

• Decision Trees (DTs) provide a competitive and efficient alternative for DOA estimation in signal reception.
• The proposed ML-based method significantly outperforms the MUSIC algorithm in accuracy and speed.
• This study highlights the potential of ML for enhancing antenna array system performance.