Improved Weighted K Nearest Neighbor Algorithm for Indoor Visible Light Fingerprint Positioning System

Aiming at the problem that the Euclidean distance in the weighted K nearest neighbor (WKNN) algorithm can not effectively represent the actual distance relationship between measurement points in the indoor visible light fingerprint positioning system, an improved WKNN algorithm based on weighted Euclidean distance measurement is proposed in this paper. The algorithm assigns different weighting coefficients to different signal strength differences according to the attenuation characteristics of the received signal strength varying with the actual distance. The simulation results show that under the same environmental conditions, compared with the WKNN algorithm using European distance measurement and Manhattan distance measurement, the average positioning error of the improved algorithm is reduced by 37. 5% and 34. 3%, respectively.