Aiming at the problems of long time and low positioning accuracy of indoor positioning algorithm, an indoor positioning algorithm based on improved LightGBM algorithm is proposed. The algorithm first preprocesses the data in the fingerprint database, and removes outliers and outliers through KNN algorithm to reduce environmental noise interference and improve data reliability. Next, divide the sample set into training set and test set, and use LightGBM algorithm to model. At the same time, genetic algorithm is used to adjust the parameters in LightGBM algorithm, and the optimal parameters are found according to the fitness function to obtain the LightGBM+GA coordinate prediction model. Finally, a prediction model is established according to the optimized parameters to realize coordinate prediction. The experimental results show that the algorithm improves the accuracy of WiFi positioning by 0.1m compared with XGBoost algorithm and 0.19m compared with GBDT algorithm. In terms of positioning time, LightGBM+GA algorithm is 5.10s faster than GBDT algorithm and 5.97s faster than XGBoost algorithm, which has good practicability.[ ]