Abstract: The large scale optimization problem of WCDNA wireless networks base station planning is studied by means of the basic principle of rolling window optimization. A WCDMA base station planning method based on rolling windows is presented, where the global optimization problem is decomposed into small scale optimization problems in rolling windows. An effective shifting strategy is designed to move the windows in two-dimensional space, and the inter-interference in WCDMA networks is predicted for each rolling window. It is proved that our method keeps the global objective of the original problem non-increasing during optimization. In simulation, three different shifting strategies are tested, and the results are analyzed. This paper shows the importance of shifting strategy for rolling windows in two-dimensional space.