Indoor WiFi initial position is challenging especially when deployed over wireless device with limited system resource, of course,the initial positioning of the mobile robot is a significant task. Although GPS(Global Positioning System) can give approximate position of the mobile users, it is usually limited indoor due to the degradation of signals by the building structures. While various alternative WiFi initial position techniques have been proposed to indoor uses, accurate results are hard to achieve due to the instability nature of wireless signal. In this paper, based on the ROS(Robot Operating System) platform,we design a WiFi indoor initialize positioning system by triangulation algorithm. We will show several experiments made in our university building to test the system. The major contributions to the presented work are that the initialize positioning error goes down and it has fast convergence compared with the method of global_localization service. The test results show that the WiFi indoor initialize position system combined with AMCL( Adaptive Monte Carlo Localization) algorithm can be accurately positioned and has high commercial value.