The Open Automation and Control Systems Journal

2014, 6 : 1240-1245
Published online 2014 December 31. DOI: 10.2174/1874444301406011240
Publisher ID: TOAUTOCJ-6-1240

Study on Network Complexity Based on Clustering Algorithm

Xinlei Li
Henan Normal University, Henan Xinxiang 453007, China.

ABSTRACT

By introducing concept of complex system into computer network system, the computer network can be regarded as a complex system with huge volume of data and information contents, including local area network (LAN), Internet and wide area network (WAN), etc. The relationships of peak and valley of network and the conflict and competition are found out through tracking network. then the relationship among each member is optimized so as to achieve the reasonable operation of complex system. In the case of classifying the researches on network complexity, the computer system is divided into TCP/IP system, ISO/OSI reference model, and so on, or it is classified in accordance with competitive characteristics of different networks. Based on clustering algorithm, this study presents the network complexity in terms of network flow, and compares the accuracy of different clustering algorithms.

Keywords:

Network complexity, Clustering algorithm, EM clustering algorithm.