基于同步时间Petri网的细粒度多媒体同步模型研究

多媒体同步是分布式多媒体信息系统的关键问题。在分析和比较几种已有的典型的多媒体同步模型的基础上．本文提出了一种新的模型：同步时间Petri网(简记为STPN)。STPN不仅能完整、准确地对媒体流内和媒体流间的时间层次行为进行形式化描述，而且解决了时间流Petri网中存在的死托肯问题。特别地，通过引入抑止弧，STPN有效地克服了随机延迟、阻塞、丢包等因素的不良影响。尤其当丢失率较高时．STPN更能体现其优越性和鲁棒性。

Fine-Grained Distributed Multimedia Synchronization Model Based on Synchronization Time Petri Nets

Multimedia synchronization is the key of distributed multimedia information systems. Based on the analysis and comparison of several typical multimedia synchronization models, this paper presents a new model: Synchronization Time Petri Nets- STPN not only allows the temporal behavior of both intra-media and inter-media to be described fully, accurately, and formally but also solves the problem that time stream Petri net has dead tokens. In particular, STPN introduces inhibitor arcs to handle the problem of synchronization failure because of fire interruption caused by random delay, blocking, or lost packets. Furthermore, STPN is robust for loss of objects in some extent and can effectively avoid long time disparity of media synchronization, especially, in case of high loss rate of media objects.