摘要
为了更真实地模拟网络行为,提出了一种基于无尺度网络的数据包传输元胞自动机模型,模型中通过增加缓存队列大小和提高中心节点的处理速度来减少丢包数,提高网络性能。仿真结果表明,在元胞交互规则的作用下,数据包发送速率,缓存队列大小和丢包数在时间序列上都具有自相似特性,自相似程度呈现负相关;它们在时间序列上是长程相关的,在功率谱图中都具有幂律分布特性,在高频段呈现出白噪声特性。
To simulate the network behavior more close to the reality,this paper brought a cellular automata model for packet transmission under a kind of scale-free network. In this model,according to increase the size of buffer queue or promoting the processing speed of the center node could reduce the discard packets and promote the network performance. The simulation results show that there is the self-similarity property in the rate of packet,the size of buffer queue and the number of packet loss,and the degree of its self-similarity presents negative correlation. They are long-range correlation on the time series,and their high-frequency section presents white noise property in the power spectrum,at the same time there is the power-law property in power spectrum.
出处
《计算机应用研究》
CSCD
北大核心
2010年第12期4686-4689,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60873079)
重庆市自然科学基金计划重点项目(CSTC2009BA2089)
关键词
无尺度网络
元胞自动机
丢包数
自相似性
功率谱
幂律特性
scale-free network
cellular automata
packet loss
self-similarity
power spectrum
power-law
