| As the rapid development of Internet technology,Internet applications are becoming increasingly popular on a global scale.In the current era,i nformation technology is advancing rapidly,and information technology is changing the way of life in society.The rapid development of the Internet provides a broad platform for the application of new technologies.The increase in streaming media technology and access network bandwidth has promoted the transmission of multimedia content on the Internet.This new service is for Internet TV and IP-based audio / video transmission.Provided the possibility,Vo IP(Voice over Internet Protocol)technology came into being.Vo IP is called IP phone or Internet phone,which is a communication system based on the Internet for voice interaction.With the lower cost of voice services and the advantages of promoting the utilization of network resources,Vo IP systems have developed rapidly around the world.The main limitation of this technology is the need to ensure a certain level of service quality based on stable transmission conditions,and the IP network was originally designed for data transmission,which uses a connectionless "best effort" transmission Way,did not provide any quality of service.In the Vo IP system,due to the technical characteristics of packet switching and the unstable network environment(such as network congestion),packet loss and delay are very common,resulting in a rapid deterioration of voice communication quality.Therefore,how to guarantee the quality of service of real-time voice communication in the presence of packet loss is a key issue to be solved urgently by Vo IP technology.A key technology of the Vo IP system is the voice codec technology,which directly determines the performance of the Vo IP system.Different from the traditional voice communication compression sensing coding and decoding method,this paper proposes an anti-packet loss algorithm based on the encoding end and decoding end under the theoretical framework of compressed sensing(CS),whose main contributions are as follows: The end constructs a sparse binary perception matrix that satisfies low coding complexity and high perception ability,performs overall linear resampling of the speech signal,so that each sampling point contains the overall information of the signal,and then packages it according to the TCP / IP standard through the Vo IP transmission model Send;2)In the case of random loss of important information,the decoder can use a compressed sensing reconstruction algorithm based on graph theory to use only the remaining data to recover the high-quality audio signal,so as to achieve the purpose of anti-packet loss,which is real-time for the mobile Internet Voice communication provides a new architectural approach.Secondly,the mathematical model of compression-sensing encoding and decoding is introduced,and a simulation example is carried out.In the ran dom packet loss environment of the network,an improved voice packet loss recovery algorithm based on compressed sensing is proposed in this paper,and optimizes THCHS30,Librispeech,and Common Voice Data-sets three data sets.Then compare with standard anti-packet loss algorithm based on compressed sensing,SILK codec algorithm with in-band FEC,and SILK codec algorithm without in-band FEC.Experimental results show that the speech signal quality reconstructed by the improved compressed packet loss recove ry algorithm(PLRCS)based on compressed sensing at a certain packet loss rate is better,and provides users with a good listening experience,thus proving that this method is more traditional The method has the performance advantages and enforceability.Finally,with the help of SIP protocol,a class of Vo IP system based on Android application framework is designed,its functions are implemented,and the server registration module,short message module,call service module and so on are tested.Then,capture the data packets of the RTP stream,and analyze the voice parameters such as delay,jitter,and packet loss rate in detail. |
PDF Full Text Request