This paper introduces a new SAS monitoring system. The system utilizes a self-designed app invoked by mobile phones to collect the data of users' snore. Then it transmits the data via home wireless network by means of ftp and finally stores the data in a PC. Moreover, the neural net algorithm and voice-recognition technology have been inserted into core algorithm of the system to identify voice and snore, which can implement the diagnosis of the SAS by combining the analyses of symptoms. The system has a higher disease classification rate than a traditional SAS monitoring system.