Performance Analysis of Softly Inhibitive Cellular Networks Using Poisson Mean-Field Approximation

The Strauss point process is an essential and softly inhibitive model for describing random wireless networks. In this letter, we introduce the mean-field approximation for the Strauss point process achieved through minimizing the Kullback-Leibler divergence between its pseudolikelihood and a corresponding Poisson distribution. Through simulations, we also show that the distribution of empty space distance for the Strauss point process can be accurately modeled by the Weibull distribution, with its scale and shape parameters efficiently determined using regression formulas that rely on the Strauss model parameters. These findings enable us to derive analytically tractable performance metrics, such as the probability of coverage and average data rate, which are validated through Monte Carlo simulations.