Location-based services have become an indispensable basic service in people's daily life and work. However, in the indoor environment, there is still no widely used positioning technology. Therefore, a large number of researchers have been engaged in the research of indoor positioning technology. Because of its wide deployment and low cost, WLAN has received the most attention, so we concentrate on the indoor positioning technology based on WLAN.A new algorithm is proposed to improve the K-nearest neighbor algorithm. The advantages of the proposed algorithm are analyzed theoretically and experimentally verified.We design an indoor positioning system based on the algorithm above.As for the heterogeneity problem caused by smartphones in the system,Signal Strength Difference (SSD) is used to replace the Received Signal Strength (RSS) to calculate the location. The reasons for using SSD are analyzed and verified through experiments. Dynamic conditions are studied because the speed of improved K-Nearest Neighbor algorithm can't satisfy the requirement in dynamic scene. We take advantage of inertial sensor of the smart phone. Using dead reckoning to measure the position change of the user during the walking and do experiment under the dynamic scene.Finally, the implementation of the system are introduced. The client end includes data collection APP and positioning APP base on Android System. The server end includes location fingerprint database, an API module and an algorithm module. |