Block Kriging Based Mobile User Distribution Estimation for Antenna Tilt Optimization

Operators of mobile carriers need to cope with Quality of Service (QoS) degradation caused by the recent increase in mobile data traffic. Tuning parameters of base stations is effective to improve QoS, but selecting appropriate parameters is too hard because various factors such as base station attributes and surrounding environments must be taken into account. Antenna tilt optimization methods have been studied as approaches to improve QoS. Since these methods often utilize location data of mobile users to optimize antenna tilts, estimating mobile user distribution accuracy is important to improve the effectiveness of these methods. Most methods utilize Minimization of Drive Test (MDT) data for obtaining the distribution or assume that this location data is known. However, estimating mobile user distribution by MDT is hard because user data collected by MDT is limited. We thus propose Cellular Block Kriging (CBK): a mobile user distribution estimation method for antenna tilt optimization to improve QoS. Specifically, CBK is an extension of block kriging, a widely used spatial interpolation method, to cellular networks. We evaluate the effectiveness of CBK through simulations.