Network Complexity: Cross-Layer Models and Characteristics

Many fundamental important questions regarding Internet processes and protocols cross-layer interaction are still remained unexplored. One of the questions is network complexity as a character of network behavior. Today complexity has three distinct and almost opposite meanings. In one case, complexity is synonymous with fluctuation in traffic dynamics and the lack of predictable structure. The second sense of complexity refers instead to how structured, hierarchical, and sophisticated a network process is. That is, "complexity" is an indicator of how many layers of order is embedded in a process. The third sense is Kolmogorov Complexity or measure of compactness of algorithmic code that can be used to modeling process. This paper considers complexity as a on of fundamental property that can be applied to describe cross-layer interaction, fractal characteristics and invariant properties caused by elastic dynamics of the virtual TCP connection and statistical distribution of retransmission.