Research On The Optimization Of The Wireless Transmission Technology For Layered Source Coding

Recently, with the rapid development of wireless communication and computer networks, wireless multimedia service has become one of the most popular applications in the current networks. Wireless multimedia service, especially the video service, has a higher requirement in transmission rate, real-time and transmission error rate. Therefore, with the increasing demand for wireless multimedia service, it will be a great challenge for the wireless communication technology, since massive data have to be sent on the unreliable wireless channel with narrow bandwidth. Thus, the optimization of the wireless transmission technology for layered source coding has become one of the major research topics.This dissertation focuses on the optimization problem of the wireless transmission technology for layered source coding in both the point-to-multipoint broadcasting and cooperative relaying scenarios. The major work and contributions of this dissertation consist in:1. The performance analysis of the Point-to-Multipoint wireless broadcast system with LSSC (Layered Source coding with Superposition Coding) transmission model is presented. And using continuum superposition coding with source coding of infinite layers, the upper bound of the LSSC transmission approach is given by the derivation of the optimal power distribution function among the superposed layers. We will show the advantages of the proposed LSSC broadcasting approach over the conventional broadcasting (CB) approach from several aspects. The analytical results show that, by promoting the experience of stronger users (users with better channel conditions) while the performance of the weaker users (users with worse channel conditions) is maintained, the LSSC approach can greatly increase the average system performance of the broadcast system.2. Relaying cooperative communication is very popular recently since the cooperative diversity is regarded as an effect way to combat channel fading. However, in the convention cooperative protocols such as Decode-and-Forward (DF) and Amplify-and-Forward (AF), cooperative diversity is obtained at the expense of multiplexing loss since the performance is limited by the half-duplex relay model. In this dissertation, using superposition coding, two advanced cooperative protocols, i.e., Superposition Coding based DF (SCDF) and Superposition Coding based AF (SCAF), are proposed based on the conventional AF and DF protocols. The Diversity and Multiplexing Tradeoff (DMT) performances of the proposed protocols are derived under different channel assumptions. Analytical results show that second order diversity can be obtained in both SCDF and SCAF protocols as in the conventional DF and AF protocols, while only first order diversity can be obtained in the Direct Transmission (DT) protocol. In addition, the multiplexing gain in the conventional DF and AF protocol can be recovered from 1/2 to N/N+1 in the proposed SCDF and SCAF protocols. Obviously, with the increase of slot numberN, the multiplexing gain can approach to 1, the same as the Direct Transmission (DT) protocols. Therefore, from the viewpoint of DMT, the performances of the conventional protocols are greatly improved by the proposed two protocols.3. Considering the characteristics of both source coding technology and transmission technology, the optimization problem for the transmission of different source coding methods in the cooperative relaying networks is analyzed. The end-to-end distortion exponent will be derived with various source coding schemes such as single layer coding, discrete finite layers source coding and continuum infinite layers source coding. The optimum DMT allocation algorithm in the proposed SCDF and SCAF transmission protocols will also be presented. Analytical results show that, comparing with the conventional DF/AF transmission approach, the proposed SCDF/SCAF transmission approach can always achieve higher distortion exponent for all bandwidth ratios. And comparing with the DT transmission approach, the proposed SCDF/SCAF transmission approach can outperform the DT approach greatly under larger bandwidth ratio, leading to a maximum distortion exponent of 2.4. Considering the real scenario, the transmission model of two layers source coding with MDF (Modified DF) is proposed. The optimum DMT allocation algorithm to achieve the maximum distortion exponent is also derived. Analytical results show that, under smaller bandwidth ratio, the proposed MDF approach will outperform all the conventional cooperative transmission approaches including DF/AF and SCDF/SCAF, since the source node can always keep transmission in both slots in MDF rather than only one slot in the conventional cooperative protocols. Therefore, the MDF approach can achieve the same performance as the DT approach at smaller bandwidth ratio. On the other hand, at high bandwidth ratio, MDF can also achieve the second order of distortion exponent as in DF/AF and SCDF/SCAF. And in the conventional DT approach, only first order distortion exponent can be achieved at larger bandwidth ratio. Moreover, the performance of the broadcast system based on MDF is also discussed. Simulation results show that, with MDF transmission, the performance of both the remote and near user in the broadcast system can be greatly improved from the view point of end-to-end distortion.