Abstract: Spectrum resource has become increasingly scarce in recent years, and research on cognitive radio networks has attracted much attention. Video streaming has the feature of delay-sensitive and high bandwidth demands, so routing scheme design is challenging. This paper proposes a cluster-based routing method for video streaming in dynamic spectrum environment. Considering the spectrum heterogeneity and dynamic allocation feature, this paper analyzes the optimal scale of clustering to minimize the distortion in transmission, and proposes an importance based cluster-head selection strategy, a spectrum aware multimedia routing algorithm is designed. Simulation experiments show that, the algorithm obtains ideal PSNR for multimedia sources through the cluster-based method, can provide reliable visual effect and Quality of Service （QoS） guarantee.

Key words: cognitive radio network, video streaming, clustering, Quality of Service（QoS）, group of picture

摘要： 随着可用频谱的日益稀缺和无线业务量剧增，认知无线网相关的研究逐渐成为热点；视频流具有延迟敏感、高带宽需求等特征，动态频谱分配环境下的路由设计和服务质量具有挑战性。针对该问题，为视频业务提出一种基于分簇的路由方法。考虑到频谱的异构性和动态分配特征，对网络节点的分簇规模进行分析和优化，获取能够降低传输失真的分簇方案，并提出一个基于重要度的簇头选举规则；在分簇的基础上设计了与频谱选择相结合的多媒体路由算法。仿真结果表明，该文提出的算法能够为认知环境下的视频传输提供可靠的峰值信噪比、视觉效果和服务质量保障。

关键词: 认知无线网, 视频流, 分簇, 服务质量, 图像组