Differentiate Chaos Characters from Miniature Network Flow Data Sampled in Low Resolution
低分辨率小规模网络流量数据的混沌特性鉴别

In this paper, chaos characters of miniature network flow data sampled in low resolution are differentiated, using methods of nonlinear time series analysis. Firstly, a smoothing method is given for network flow data. Then, largest Lyapunov exponent of flow data is computed, and noise data with the same characters are distinguished from network flow data. From different points of view, it proofs that network flow is a chaos system. These work provide the foundation for studying behavior characters of network flow using chaotic theory.