The Research And Design And Implement Of Mixing Technology In VOIP Conference System

The improvement of voice over internet protocol(VOIP)technology depends mainly on the development of two areas of technology for the increasing of network bandwidth and the applications of VOIP communication protocol,such as,H.323,MGCP,RTP,RTCP and SIP protocols,and so on.On the other hand,the development of multimedia technology,i.e.,the performances have been continuous improving of the technologies of digital signal processing(DSP),voice,image,decompression,and so on.The media gateway is a key factor to deal with the multimedia in the VOIP system,and its performance determines the performance of VOIP system.In the media gateway,the audio module is a key component in the teleconferencing system.Its performances determine the quality of voice in the conference system.Meanwhile the quality is one of the important parts of judging the quality of the conference system.Traditional conferences systems,firstly,it will select a conferences chairman,the participants must apply to the chairman of the meeting to obtain the right of speaking,at the same time.One person is allowed to speak at one time,only the speaker released the right of the speech,and another participant may apply to the chairman of the conference,repeat this process.Such the way to greatly limit the exchange among participants,did not take full advantage of the features of shortcut and convenience of the network technology.And the use of remix has greatly enhanced the participants' sense of on-site meeting,which allowing more participants to speak and convey the speaker's voice message to other participants in the meeting simultaneously.This paper briefly introduces the reasons and the basic working principle for the rapid development of VOIP,and corresponding protocols,coding technology,transmission technology,jitter buffer processing,quiet suppression,and echo cancellation technology,and so on.According to the requirement of product conference function,the paper introduces the mixing algorithms of commonly teleconference system and analyzes the characteristics of the ones.In order to ensure that the mixing module does not increase the time-delay of the algorithm with the increase of the number of participants,the paper designs and realizes the mixing module based on programmable logic device.The results show that no matter what kind of mixing algorithm will be adopted,when the participants became larger,the algorithm will produce a lot of noise by the overflow of amplitude of voice.So,this paper studies mute detection of the voice.Because of the different requirements of silent data detection and terminal mute detection in the mixing module,the classical short-time energy spectrum algorithm is used to distinguish speech and background noise.At the same time,this paper presents a mixing algorithm using energy spectrum as a control strategy.That is,the use of short-term energy spectrum parameters to distinguish voice and noise.The parameters are also used as a judgment for controlling whether the participant's voice is involved in the mixing,we select the first three major voice to participate in the mixing,while the other participants in the mixing of the voice set to mute,thus greatly reduces the noise by the voice overflow.The actual test of the algorithm show that the algorithm satisfies the application requirements of the customers,the correctness,feasibility,and reliability are verified by the algorithm.