| 摘要: |
| 研究了自相似网络通信量的统计复用策略。首先讨论了在采用FIFO(First In First Out)服务策略且缓存趋于无限大时，满足一定溢出概率要求的统计复用带宽要求，然后对这种情况进行了仿真分析。研究表明网络通信量的自相似特性对复用增益有较大的影响，并且自相似性越强、复用通信量数量越大，所取得的复用增益越大，这对于网络设计是很重要的。 |
| 关键词: 通信网 自相似网络 复用策略 溢出概率 无限缓存 |
| DOI：10.3969/j.issn.1001-893X. |
|
| 基金项目: |
|
| Multiplexing Scheme for Self-Similar Network Traffic with Infinite Buffer |
|
| () |
| Abstract: |
| The statistical multiplexing scheme for self-similar network traffic is investigated. First, with the FIFO (First In First Out) service discipline and pre-determined buffer overflow probability, the requirement of multiplexing bandwidth is obtained under the condition of asymptotic infinite buffer. Results show that the self-similarity of network traffic has a significant influence on the statistical multiplexing gain. |
| Key words: Communication network,Self-similar network,Multiplexing scheme,Overflow probability,Infinite buffer |
