规则全光分组网络中偏射路由算法的分析

就在ShuffleNet和Manhattan Street Network两种规则网络中使用偏射路由算法后的网络性能以及允许一个时隙插入多个数据包对该算法的影响进行了分析。结果说明，偏射路由算法不仅能使网络得到较高的性能，而且发挥了网状网具有迂回路由的能力。当采用允许插入多个数据包的策略时，网络的吞吐量和平均跳转次数都有小幅度的增加。

Analysis of deflection routing in regular optical packet switching network

We focus on analyzing the performance of the ShuffleNet and Manhattan Street Network using deflection routing and the effects on it when two packets can be inserted into node. The results show that deflection routing is very effective. It not only makes network perform very well in the lack of buffers but also mines the ability of mesh network which has bypass routes. When new tactic that two packets can be inserted is used, the throughput of network increases while the expected hops also increases.