Abstract: Aiming at the diversity of bearer service for multimedia sensor networks and QoS requirements for different services（e.g.video/image services，audio services and data services），a business-driven QoS routing model is abstracted for multimedia sensor networks and a Service-Driver Routing algorithm（SDR） is proposed based on ant colony algorithm and chaos theory.The SDR algorithm selects corresponding routing strategy according to different QoS requirements for different services，to reach the targets of using the network resources fully and enhancing the network load business performance.Finally，the routing algorithm is simulated using NS2.Compared with other methods，the simulation results show that the proposed algorithm can better satisfy the QoS requirements for differentiated services.

Key words: service-driven, multimedia sensor networks, chaos ant colony optimization, Quality of Service（QoS） routing

摘要： 针对多媒体传感器网络承载业务（视频图像业务、音频业务以及数据业务）的差异性和不同业务对QoS的差异性，基于蚁群算法和混沌理论，抽象业务驱动的QoS路由模型，提出了一种业务驱动的路由算法SDR（Service-Driver Routing）。依据各类业务对QoS的差异性选择相应路由策略，达到充分利用网络资源、提高网络承载业务性能的目标。采用NS2对该算法进行实验仿真，实验结果表明，与其他路由算法相比，该算法可提供区分业务的QoS保障。

关键词: 业务驱动, 多媒体传感器网络, 混沌蚁群算法, 服务质量（QoS）路由