提出了光突发交换网络支持服务质量的一种新方案.在此方案中,核心节点根据到达的数据突发的优先级别,将每一出口链路的数据信道进行分组.每一组别的数据信道数主要取决于相应优先级别的数据流量.通常情况下,属于某优先级别的数据突发会被调度在相应组别的数据信道上.在高优先级别的数据突发未能预留在其相应组别的数据信道上时,可尝试为其在低组别的数据信道上预留,甚至抢占低组别内已经被预留的信道;低优先级别的数据突发不能被预留在高组别的数据信道上.另外,也考察了抢占数据突发长度与被抢占数据突发长度之间的关系.

A new approach in optical burst switching(OBS) networks with QoS support is presented in this paper. In the approach,the data channels of an outgoing link at a core node are divided into multiple groups,with each group corresponding to a service class. The number of data channels in each group is mainly determined by data traffic. In general,a data burst(DB) can be sent on a data channel reserved by its burst head packet(BHP) only in its own group. Upon failing to reserve any bandwidth in its own group,the BHP tries to re-reserve even preempted bandwidth on a data channel in a lower-priority group. A lower-priority BHP can't reserve bandwidth on any data channel in a higher-priority group. In addition,the reasonable relation between the preempting DB length and the preempted DB length is also investigated in this paper.

Supported by the National Natural Science Foundation of China under Grant No.90304004(国家自然科学基金);the Base Programs for Application of Chongqing Science and Technology Committee of China under Grant Nos.8060,8061(重庆市科委应用基础项目);the Programs for Science & Tech