Objective In order to eliminate the influence of sleep posture changes on respiration volume monitoring, we propose a novel method based on dual-channel bioelectrical impedance technology combined with data fusion algorithm to extract the respiration volume signal from the left and right lung respiration impedance signals for respiration volume monitoring.Methods In this paper, a sleep respiration volume monitoring system based on dual-channel bioelectrical impedance technology is built, and the system acquires respiration signals by the multi-channel physiological signal recorder MP150 (BIOPAC, USA) .Ten subjects with normal breathing in turn, keep three sleep postures:dorsal decubitus, left lateral decubitus and right lateral decubitus, andrespiratory airflow signals and left lung, right lung and thorax respiratory impedance signals are collected.We set the respiratory airflow signals as the gold standard signals, and compare the left and right lung fusion signals obtained by the factor analysis method and the respiratory impedance signals of the single channel thoracic cavity.Results Compared with single channel impedance method, the difference between the left and right lung respiration signals and the gold standard signal value was smaller, and has a better consistency.Conclusions The method based on dual-channel impedance technology combined with factor analysis-based data fusion algorithm to extract respiration volume is a novel method of sleep respiration volume monitoring with high credibility.