Channel state information fingerprint positioning method based on weighted k neighbor

The invention discloses a channel state information fingerprint positioning method based on weighted k-nearest neighbor, and belongs to the technical field of indoor positioning. According to the method, an existing CSI fingerprint positioning system is improved, and the method comprises the steps that a neighbor point recall scheme based on prediction of a real spatial distance is adopted to improve neighbor point recall quality, and an inhaul cable regression model is preferably selected as a neighbor point recall model; adopting a neighbor weighting scheme based on a neural network fitting weighting coefficient to improve neighbor weight setting, and adopting a form of adding a prior linear equation to improve a convergence problem caused by insufficient input and output correlation of the neural network; and during online positioning, selecting k neighbor points closest to a point to be measured by using the trained neighbor point recall model, predicting weighting coefficients of the k neighbor points by