Key Technologies Of Encoding And Transmission In Mobile Streaming Media

With the continuous development of wireless networks and wireless terminal equipment,streaming media business has become an important direction of the current and future in wireless communications domain.However,the source streaming data is huge.And the limited bandwidth makes the wireless channel affected to a large extent by noise and scale fading.Moreover,the mobile terminal is moving,causes the radio signal uneven.All this above result in data packet loss,delay and other quality of service reduced.Therefore,to compression coding of the source,stable transmission of the channel and the establishment of a complete simulation platform can provide help for the fast and accurate signal transmission and reconstruction.This thesis studies the excisting problems in technology of source sampling,encoding and transmission of streaming media.Firstly,study the compression algorithm of video data in the sampling field.Since the main source of streaming media-video data has the characteristics of large volumes of data.After sampling under the traditional sampling theorem Nyquist,the data often become a much larger sample.While,the latter part of the video encoder will try to eliminate the redundancy as much as possible.The process results in a huge waste of resources.To solve this problem,compressed sensing algorithm is introducted.And according to the characteristic of inter redundancy among the video frames,video compressed sensing based on principal component analysis(PCA)is designed.So the signal sampling and compression can be processed simultaneously,so that the sampling frequency is greatly reduced.In the process of PCA,the covariance matrix of video signal is big enough to make the computer system overflow.Then video compressed sensing with GHA is proposed.The algorithm determines the sparsity,which also means the number of principal components,by meannent matrix in compressed sensing.Then the computing problem of the huge complete covariance matrix is solved and the system efficiency is improved.Secondly,study and improve the motion estimation algorithm in video coding technology.In this thesis,the H.264 standard is a research platform.And we do in-depth analysis of the codec framework and key technologies in the standard.Motion estimation algorithm is the most important part in the encoder.After analyzing the advantages and disadvantages of the typical block matching algorithm,an improved adaptive road pattern search(IARPS)algorithm is put forward.For the searching point is not accurate enough,spatial and temporal adjacent block are both used as the reference macroblocks of initial prediction.We also use variable-step search method.And for search template,instead of the traditional idea of template type close to the circular,we choose cross-based search template based on statistical data.So the amount of the search is reduced,and the search precision is increased.Thereby we can reduce the search points of the motion estimation algorithm,and increase the search efficiency.It is also equivalent to improve the coding efficiency.Then,OFDM transmission system of streaming media is carried out.For multi-path effects of wireless network will cause frequency-selective fading,inter-symbol interference of the received signal makes communication performance declines.A transmission system based on orthogonal frequency division multiplexing(OFDM)is established.Not only solve the problems caused by frequency selective fading,but also improve the spectrum utilization efficiency.To achieve the OFDM transmission of streaming media,H.264 NALU transmission unit is first proposed.And the number of OFDM subcarriersand is determined by the bits in the NALU adaptively to create OFDM transmission system of streaming media.To avoid the poor immunity problems of traditional OFDM system based on FFT,orthogonal wavelet packet function is used to build OFDM systems.With the good mutual orthogonality and self orthogonality,wavelet packet function can better combat channel delay,interference and inter-symbol interference between subcarriers.Finally,adaptive error control algorithm for QOS protection mechanism in streaming media is studied.An adaptive FEC algorithm with FEC redundant packets dynamically changing with the network load and the state of wireless channel is designed.The existing algorithm sets the network load as a primary factor artificially,while the wireless channel status as a factor of secondary importance.Under the problem of inadequate application of the channel state,the algorithm has been improved,increasing the importance of wireless channel state in the calculation of FEC redundant packets,and enhancing the channel utilization.Since FEC redundancy package changes too fast at the threshold point for the existing theoretical calculation resulting from the use of a linear function,an adaptive FEC control algorithm with nonlinear threshold is designed based on the cosine function.The traditional streaming media encoding and transmission platforms have heterogeneity.Here Evalvid architecture and NS2 simulation platform are adopted to establish a complete simulation platform for streaming media with a set of source coding-channel transmission-signal reconstruction-QoS analysis.The platform is also used to test the error control performance of the adaptive algorithm under different error model channels.And quality service of video transmission in streaming media simulation platform is analyzed.Results of this study have important significance and engineering value.The simulation platform of mobile streaming media built can provide a strong theoretical simulation and analysis tools for mobile streaming media encoding and transmission process.Part of the research results can be applied to the field of video encoding compression,signal processing and signal transmission.