A quantitative real time data analysis in vehicular speech environment with varying SNR