论文部分内容阅读
利用相变的概念解释复杂计算机网络中广泛存在的幂律现象 ,并未考虑到网络这一复杂系统中节点之间的相互关联作用 .通过一种计算机网络模型 ,对网络中的相变和整体关联行为进行了分析和比较 .研究表明 ,随着负载的增加 ,网络节点行为的空间和时间的长程相关性逐渐增强 ,整体关联行为逐渐显现 ;在临界状态 ,网络负载变化的功率谱呈现 1 f2 的特点 ,而节点排队长度变化的功率谱服从 1 fα 分布 ,其中α≤ 2 .网络节点行为在时间和空间上的关联进一步得到说明 .
The use of the concept of phase transition to explain power law phenomena that exist widely in complex computer networks does not take into account the interconnection of nodes in a complex network.According to a computer network model, The research shows that with the increase of load, the long-range correlation between space and time of network node behavior gradually increases, and the overall correlation behavior gradually appears. In the critical state, the power spectrum of network load changes presents 1 f2 , While the power spectrum of node queuing length varies with 1 fα distribution, where α ≤ 2. The temporal and spatial correlation of network node behavior is further illustrated.