ESGWKNN: Research on Indoor Localization Algorithm Based on Wi-Fi

Abstract Aiming at the problem that the accuracy of the traditional WiFi-based K-Nearest Neighbor algorithm (KNN) indoor positioning algorithm cannot meet the requirements of precise positioning, this paper proposes a K-nearest neighbor indoor positioning algorithm based on location range limitation, namely the algorithm combining echo state network and Gaussian weighted K-nearest neighbors (ESGWKNN). The ESGWKNN algorithm first uses the echo state network to estimate the range to obtain the local spatial range information, and then uses the GWKNN algorithm to accurately locate within the local spatial range to reduce the positioning accuracy problem caused by excessive space. The experimental results show that compared with the traditional KNN algorithm, the ESGWKNN algorithm has higher positioning accuracy.