节点文献

基于H.264的无线传输差错控制及解码器的ARM实现

Error Control for Wireless Transmission and the Decoder Implementation with ARM Based on the H.264

【作者】 桑亚林

【导师】 张正华;

【作者基本信息】 扬州大学 , 信号与信息处理, 2009, 硕士

【摘要】 信息化社会的到来以及IP技术的兴起,正深刻的改变着电信网络的面貌以及未来技术发展的走向。无线通信技术的发展为实现数字化社区提供了有力的保证。而视频通信则成为多媒体业务的核心。如何在环境恶劣的无线环境中,实时传输高质量的视频面临着巨大的挑战,因此这也成为人们的研究热点。对于无线移动信道来说,网络的可用带宽是有限的。由于多径﹑衰落﹑时延扩展﹑噪声影响和信道干扰等原因,无线移动通信不仅具有带宽波动的特点,而且信道误码率高,经常会出现连续的﹑突发性的传输错误。无线信道可用带宽与传输速率的时变特性,使得传输的可靠性大为降低。视频播放具有严格的实时性要求,这就要求网络为视频的传输提供足够的带宽﹑有保障的延时和误码率。为了获得可接受的重建视频质量,视频传输至少需要28Kbps左右的带宽。而且视频传输对时延非常敏感。然而无线移动网络却无法提供可靠的服务质量。基于无线视频通信面临的挑战,本文在对新一代视频编码国际标准H.264/AVC研究的基础上,主要在提高其编码效率和H.264的无线传输抗误码性能,以及如何在嵌入式环境下实现H.264解码器进行了研究。结合低码率和帧内刷新,提出一种针对感兴趣区的可变帧内刷新方法。实验表明该方法可以使用较少的码率对感兴趣区域进行更好的错误控制,以提高区域图像质量,同时能根据感兴趣区及信道的状况自动调整宏块刷新数量,充分利用有限的码率。为了有效的平衡编码效率和抗误码能力的之间的矛盾,笔者提出了一种自适应FMO(Flexible Macroblock Order)编码方法,可根据图像的复杂度自适应地选择编码所需的FMO模式。仿真结果表明这种FMO编码方式完全可行,且在运动复杂度频繁变化时效果更加明显,完全可应用在环境恶劣的无线信道中。在对嵌入式PXA270硬件结构和X264研究的基础上,基本实现了基于H.264的嵌入式解码,在PXA270基础上进行环境的配置,定制WinCE操作系统,并编译、产生开发所用的SDK和下载内核到目标机。利用开发工具EVC实现在PC机上的实时开发和在线仿真调试,最终实现了对无差错H.264码流实时解码。

【Abstract】 The advent of information society, as well as the emergence of IP technology, are profoundly changing the face of telecommunications networks and the future direction of technology development. Wireless communication technologies provide strong pledge for the realization of the development of digital communities. Video communication is the kernel of multimedia operation. The transmission of video content over wireless networks poses big technical challenge duo to the specific characteristics of wireless channel such as bad environment, so it becomes a hotspot for the research.For the wireless channels, the available bandwidth is limited. Due to the multipath effect, multipath delay, multipath fading and some other reasons , the mobile communication has the characteristics that its bandwidth is undulate and high bit error rate, so some continuous and gusty error of transmission usually appears. The reliability of the transmission decrease because of the available bandwidth and the transmission rate are time-variant.Video display needs well timeliness, this requires the sufficient bandwidth、guaranteed time lag and the bit error rate. In order to replay the video , we need at least 28 Kbits bandwidth. Video transmission is very sensitive to the time lag. But mobile network couldn’t provide credible services.Based on the challenge of the wireless video communication, this paper dose some research on the H.264 , especially on how to well control error diffuseness and improve coding efficiency as well as how to realize the H.264 decoding on ARM platform.In wireless communication system, there are many great challenges in transforming reliable video signals. So, how to effectively control bits error becomes urgent and significant. In this method is proposed to control bits error in low bit rate wireless transmission system. This method can use less overhead bits to control bits error of the interesting areas to improve the quality of the area of the image, and can automatically adjust the number of MB( Macroblock ) to make fully use of the limited bits rate according to the location, the size , and the shape of the interesting area.In order to effective balance contradiction between coding efficiency and the anti-error code ability, the author proposed one auto-adapted FMO code method; could auto_adapted and choice FMO encoding pattern needed according to the image complexity.The simulation result indicated that this FMO encoding method is completely feasible, especially the effect is more obvious when movement complexity frequent change.It can apply in the bad wireless channel environment.Based on the research on the hardware of the PXA270 and the X264, we basically realized decoding process on ARM platform. Do some environment setting on PXA270,customize WinCE Operation System,compile,produce SDK for application exploitation and download kernel into target ship. make use of exploit tool EVC ,make application program realtime and debug on line, finally realized realtime decoding for no error H.264 code.

  • 【网络出版投稿人】 扬州大学
  • 【网络出版年期】2009年 12期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络