Scale-free Networks with Self-Similarity Degree Exponents

Ravasz et al. structured a deterministic model of a geometrically growing network to describe metabolic networks. Inspired by the model of Ravasz et al., a random model of a geometrically growing network is proposed. It is a model of copying nodes continuously and can better describe metabolic networks than the model of Ravasz et al. Analysis shows that the analytic method based on uniform distributions (i.e., Barabási-Albert method) is not suitable for the analysis of the model and the simulation process is beyond computing power owing to its geometric growth mechanism. The model can be better analyzed by the Poisson process. Results show that the model is scale-free with a self-similarity degree exponent, which is dependent on the common ratio of the growth process and similar to that of fractal networks.