摘要
本文讨论了两种不同的冲突解决方法——回退法和保持法。性能分析的模型是基于一种更实际的假设,定量分析的参数是平均请求服务时间,分析的结论得到了模拟结果的验证,即回退法比保持法好。保持法在高请求率的情况下,要多占用结点周期。
In this Paper, two difference Conflict resolution methods, backtrace and hold, are discussed. Our analytical models of performance based on a more realistic as- sumption. The quantitative analytical parameter is average request service time. The outcomes of analyses are given and are verified by simulation results. It is shown that the Dacktrace method is better than the bold method. In the hold method. the data transfer time has taken more node-cycles.
出处
《重庆邮电学院学报(自然科学版)》
1992年第1期63-74,共12页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
关键词
互连网络
性能
计算机网络
block
connect
transition probability
state