无尺度网络中的拥塞及其控制

实际网络经常承受超负荷的流量，由于网络节点自身容量或者处理速度的限制，往往导致严重的拥塞产生，使得网络具有较大的时延并且性能下降．已有研究表明，实际的通信网络具有无尺度特征．本文研究了无尺度网络模型中的拥塞现象及其控制方法，结果表明网络节点的性质和网络的无尺度特性均对拥塞现象的产生和控制有显著影响．因此，仅需要对一些最关键的节点加以控制作用，就可以得到类似控制所有节点所产生的控制效果，从而节约花费．

Congestion and decongestion in scale-free networks

Real networks often experience over_loaded traffic, congestion will occur due to the limited capacity and the process rate of network nodes, which will cause the networks the experience serious delays and performance degrade. It has been shown that real communication networks are scale_free. In this paper, the congestion and decongestion problems in scale_free networks are studied. Simulation results show that congestion and decongestion in a scale_free network are significantly influenced by both the property of the nodes and the scale_free feature of the network. Thus, the cost will be saved by only applying control strategy to some most important hub nodes, while similar control effect is achieved.