Voip Voice Quality Evaluation System In The Centralized Wlan Architecture

With the rapid development of WLAN and VoIP technology, transferring IP voice on WLAN is a hot subject of research. The concept of VoWLAN is proposed.Because of instability of wireless media, the voice quality of VoWLAN can't reach the level of legacy telephone. Research on the evaluation of voice quality base on centralized WLAN could be considered as a necessity to make improvements.By considering the features of centralized WLAN and main factors which affect VoIP audio quality, the thesis researches on the evaluation of voice quality of VoWLAN. In details, it firstly researches on the main technologies of WLAN and VoIP. It introduces centralized WLAN architecture and H.323 protocol.Secondly,it researches on applying VoIP to centralized WLAN, introduces basic concept of VoWLAN, and analyzes the problems.Then it researches on VoIP supporting technology provided by IEEE802.11 and centralized WLAN,which includes QoS implementation,radio resource management,and layer 3 roaming.Thirdly,it analyzes the factors: loss, delay, jitter and RSSI,proposes VoWLAN's QoS parameters, and introduces methods of evaluating voice quality.The emphasis is on E-Model and its modification by adding jitter parameter.Then it designs a VoWLAN voice quality evaluation system based on modified E-Model.The system identifies RTP/RTCP stream from captured UDP packets,then calculates delay and loss from the payload carried by the stream. After that,With Hurst coefficient,codec and jitter buffer, the system calculates MOS, which represents voice quality.At last,it designs simulation based on centralized WLAN architecture. The experiment uses NISTNet to simulate loss,delay and jitter in true environment.In the experiment,we used our system and PESQ to evaluate the same voice,and they got similar results.The results proved that the system can effectively evaluate VoWLAN's voice quality.