Spatial modeling of the traffic density in cellular networks

Abstract

Modeling and simulation of a cellular network typically assumes that the target area is divided into regular hexagonal cells and mobile stations (MSs) are uniformly scattered in each cell. This implies a statistically uniform distribution of traffic load over space, but in reality the spatial traffic distribution is highly non-uniform across different cells, which calls for actual spatial traffic models. In this article, we first present the analysis of traffic measurements collected from commercial cellular networks in China, and demonstrate that the spatial distribution of the traffic density (the traffic load per unit area) can be approximated by the log-normal or Weibull distribution depending on time and space. Then we propose a spatial traffic model which generates large-scale spatial traffic variations by a sum of sinusoids that captures the characteristics of log-normally distributed and spatially correlated cellular traffic. The proposed model can be directly used to generate realistic spatial traffic patterns for cellular network simulations, such as performance evaluations of network planning and load balancing.