摘要
研究无线多媒体传感网络的音频流同步问题。由于流媒体具有数据量大、实时性要求高等特征,再加上无线传感器网络的节点分布式部署、能量资源有限等特征,导致每个簇头节点要承受很大的处理任务才能满足实时性的要求。传统的无线多媒体传感网络的音频流同步方法为集中式同步处理,所有的音频流同步处理都由簇头节点承担,簇头节点能耗较大,对节点的信息丢包纠错能力弱。为了解决上述问题,提出一种全节点标识的无线多媒体传感网络音频流同步方法,在感知节点中使用特征检测法检测音频流的端点,对每个节点的信息构建音频指纹,标记同步起始点,同步任务分解在每个节点,最后在簇头节点中实现音频流同步,执行误差调整与信息纠正,降低簇头节点能耗,增强信息丢包纠错能力。仿真的结果表明,改进算法同步延时明显降低,增强了网络性能。
The synchronization problem of wireless multimedia sensor network audio stream was researched. A full node ID stream synchronization method of wireless multimedia sensor network audio was proposed, and the feature detection method was taken for audio stream terminate nodes in the sensor nodes. The audio fingerprint was construeted for each node and labeled synchronous starting point. The synchronization task was decomposed in each node. The audio stream synchronization was realized in the cluster head node. The error adjustment and correction were implemented. Then the energy consumption of cluster head nodes was reduced. The information packet loss error correcting capability was improved. Simulation result shows that the synchronous time delay of improved algorithm is reduced significantly, and the network performance is enhanced.
出处
《计算机仿真》
CSCD
北大核心
2013年第12期290-293,共4页
Computer Simulation
关键词
无线传感网络
音频同步
音频指纹
信息纠错
Wireless sensor networks （WSN）
Audio synchronization
Audio fingerprint
Information error correction
