Font Size: a A A

Study On Error-Resilience Techniques For Cellular Video Coding And Transmission

Posted on:2008-05-24Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z Y ZhangFull Text:PDF
GTID:1118360242495171Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
With the rapid progress in digital video compression and network technology, video applications over 3G or 4G cellular networks have recently received increasing attentions. Due to the low bandwidth, large bandwidth fluctuation and high burst loss rate of the wireless channels, mobile video communication over cellular networks is expected to experience burst packet losses , which will cause severe quality degradation when video bitstreaming are compressed using temporal-spatial prediction mechanism. Although the prevalent video coding standards have developed many effective error resilience techniques, there are many works are needed to develop the novel error resilience algorithms which can adapt to the cellular video communication well. In this dissertation, we focus on some key error-resilience techniques in cellular video coding and transmission. In the context of our work, cellular video is defined as the wireless video communication over the cellular mobile network. To summarize, the research includes the following aspects:Firstly, after a brief overview of the current main video coding principles and standards, the error-resilience techniques are introduced in detail. We also give the discussion of the application scenario, related protocols, major challenges and research directions for the cellular video communication.The emerged cross-layer technique is broadly adopted for wireless communication parameter optimization. Cross-layer technique breaks the restriction between layers, and thus joint optimization can be achieved through the information exchanges across different protocol layers. In chapter 2, we discuss the cross-layer based error-resilience cellular video coding techniques and propose two original schemes. The first scheme is used for the downlink cellular video transmission, where a radio-link-layer (RLL) video agent can be provided by the mobile base-station to realize the unequal link-layer retransmission (ULR). Based on the reliable transmission of relatively important frames using ULR, the encoder at the application (APP) layer can actively discard a certain number of frames according to the end-to-end feedback messages. Thus, unpredictable burst frame losses during transmission can be transformed into selective frame discarding at the encoder. Experiments results show that the proposed scheme can improve both the coding efficiency and the error resilience of the real-time cellular video communication significantly. For the condition without any video agent in the base-station, one effective error resilient video coding scheme is designed for the practical uplink cellular video delivery. Unlike the above scheme, the proposed scheme adopts a RLL agent at the sender to implement the cross-layer feedback: get the frame losses information (by prediction or time-out-frame discarding) and send it to the APP layer to adapt the encoder well to the varying channels. Since the proposed scheme does not need any additional changes on the existing mobile base-stations of the current cellular network, it provides a more feasible solution for the practical cellular video communication.Unequal loss protection is an effective error-resilience tool for the robust wireless video streaming. A critical component in the unequal-loss-protection scheme is a metric for evaluating the importance of different frames in a GOP. In the case of video streaming over 3G mobile networks, packet loss usually corresponds to whole-frame loss due to low bandwidth and small picture size, which results in high error rates and ineffective of most existing transmission-distortion-estimate models. In chapter 3, we firstly develop a recursive algorithm to compute the GOP-level transmission distortion at pixel-level precision using the pre-computed video information. Based on the study on the propagating behavior of the whole-frame-loss transmission distortion, we then propose a piecewise linear-fitting approach to achieve low-complexity transmission distortion modeling. The simulation results demonstrate that the proposed two models are accurate and robust. The proposed transmission-distortion models provide a type of fast and accurate importance-assessment tool in allocating limited channel resources optimally for the mobile streaming video.A distortion-minimized slicing scheme is proposed in chapter 4, which can be used for the unequal loss protection when delivering compressed video bitstream over packet networks. Unlike existing slicing methods, the proposed scheme uses an unequal inter-slice macroblock (MB) allocation ratio. For a given channel condition (packet loss probability and bits resource using for FEC codes), the proposed algorithm can find out the optimal packet length of the relative important slice group, which corresponds to the minimal expected end-to-end distortion. Based on that packet length, corresponding number of MBs with most importance are grouped together in a separate slice. By protecting that slice better using unequal loss protection, optimal rate-distortion performance can be achieved using the proposed video slicing scheme.Finally, we give the introduction of the CDMA2000-1X based wireless video surveillance system which is developed autonomously. In this project, DSP and CPU based embedded hardware platform is designed and implemented, and an improved congress control algorithm is also designed to adapt the system well to the varying cellular network. Unlike existing additional increase and multiplicative decreasing (AIMD) algorithm, improved congress control scheme uses a time-window based rate adjusting scheme to filter the short-time-slot bandwidth fluctuation, which is caused by the burst wireless channel errors and can result in wasting of the usable network bandwidth. Results of the experiments and practical works acceptance show that the system design is successful and the network congress control is effective.
Keywords/Search Tags:Cellular video, cross-layer feedback, unequal error protection, GOP-level transmission distortion, video slicing, congress control
PDF Full Text Request
Related items