The process of radio network optimization can be significantly improved if the real users spatial distribution is known. This knowledge can allow an optimal assignment of radio resources to particular areas of the network. This article presents 2G spatial traffic distribution analysis method and its applications in 2G and 3G radio network optimization process. Included is the algorithm of the traffic density generation and the hour-by-hour analysis of its fluctuations. Proposed is the method of cell grouping with regard to their characteristic traffic features. The article also introduces the concept of traffic character and traffic character map generation method. Finally, the following work describes practical applications of traffic density and traffic character in the 2G and 3G network optimization process.