Title: Improved binary black hole searches through better discrimination against noise transients

Abstract: The short-duration noise transients in LIGO and Virgo detectors significantly affect the search sensitivity of compact binary coalescence (CBC) signals, especially in the high mass region. In the previous work by the authors \cite{Joshi_2021}, a $\chi^2$ statistic was proposed to distinguish them from CBCs. This work is an extension where we demonstrate the improved noise-discrimination of the optimal $\chi^2$ statistic in real LIGO data. The tuning of the optimal $\chi^2$ includes accounting for the phase of the CBC signal and a well informed choice of sine-Gaussian basis vectors to discern how CBC signals and some of the most worrisome noise-transients project differently on them~\cite{sunil_2022}. We take real blip glitches (a type of short-duration noise disturbance) from the second observational (O2) run of LIGO-Hanford and LIGO-Livingston detectors. The binary black hole signals were simulated using \textsc{IMRPhenomPv2} waveform and injected into real LIGO data from the same run. We show that in comparison to the traditional $\chi^2$, the optimal $\chi^2$ improves the signal detection rate by around 4\% in a lower-mass bin ($m_1,m_2 \in [20,40]M_{\odot}$) and by more than 5\% in a higher-mass bin ($m_1,m_2 \in [60,80]M_{\odot}$), at a false alarm probability of $10^{-3}$. We find that the optimal $\chi^2$ also achieves significant improvement over the sine-Gaussian $\chi^2$.