Simultaneously search for multi-target Galactic binary gravitational waves in reduced parameter space with LMPSO-CV

We propose an innovative approach to the concurrent exploration of gravitational waves originating from Galactic binaries through the development of a new Local Maxima Particle Swarm Optimization (LMPSO) algorithm. Our methodology employs strategic Create Voids (CV) to streamline parameter space, maximizing the identification of local maxima for the ℱ-statistic even in the overlapped signals case. Subsequently, a “find-real-ℱ-statistic-analysis", which implements the astrophysical models and properties of ℱ-statistic in parameter space, is conducted to reveal Galactic binary gravitational wave signals within the dataset. Our new approach eliminates inaccuracies associated with signal subtraction contamination, which is a challenge for traditional iterative-subtraction methods when addressing low signal-to-noise ratio signals (e.g., SNR < 15). To demonstrate the efficacy of our approach, we utilize the residuals from the LISA mock data challenge (LDC1-4), where 10982 injection sources with optimal SNR > 15 have been eliminated. The LMPSO-CV method efficiently identifies 8995 signals with a 47.7% false source fraction or 3463 signals with a 26.9% false source fraction when the correlation coefficient threshold is set to 0.8.