Einstein Telescope: Detection of Binary Black Hole Gravitational Wave Signals Using Deep Learning

Expanding upon our prior work (Alhassan et al. 2022), where we evaluated the performance of single sub-detector data (SSDD) from the Einstein Telescope (ET) for binary black hole (BBH) detection using deep learning (DL). In this study, we explore the detection efficiency of BBHs using data combined from all three proposed sub-detectors of ET (TSDCD), employing five different lower frequency cutoffs (F_low): 5 Hz, 10 Hz, 15 Hz, 20 Hz, and 30 Hz, while maintaining the same match-filter Signal-to-Noise Ratio (MSNR) ranges as in our previous work: 4-5, 5-6, 6-7, 7-8, and >8. The Deep Residual Neural Network model (ResNet) was trained and evaluated for the detection of BBH gravitational wave signals using both TSDCD and SSDD. Compared to SSDD, the detection accuracy from TSDCD has shown substantial improvements, increasing from 60%, 60.5%, 84.5%, 94.5% to 78.5%, 84%, 99.5%, 100%, and 100% for sources with MSNR of 4-5, 5-6, 6-7, 7-8, and >8, respectively. In a qualitative evaluation, the ResNet model detected sources at 86.601 Gpc, with an averaged MSNR of 3.9 (averaged across the three sub-detectors) and a chirp mass of 13.632 at 5 Hz. The results demonstrate a notable accuracy improvement for lower MSNR ranges (4-5, 5-6, 6-7) by 18.5%, 24.5%, and 13%, respectively, and by 5.5% and 1.5% for higher MSNR ranges (7-8 and >8). TSDCD proves suitable for near-real-time detection and can benefit from a more powerful setup.