Fitting methods analyzing in modeling cellular networks based on stochastic geometry

The topology of base station is an important factor affecting network performance. With the increase of BSs types and density, the traditional hexagonal mesh model can no longer describe the distribution accurately. The model based on stochastic geometry, such as Poisson point process(PPP), attracts a lot of attention. This paper selected four regions of China to analyze, including urban and rural areas. Results show that the urban region can be modeled as Point cluster process(PCP) and the rural can be modeled as Strauss process(SP). This paper also used coverage probability to compare three fitting methods of Poisson cluster process. The conclusion is that the minimum contrast method is best.