Northwestern Polytechnical University
In this paper, a hybrid location algorithm for fingerprint-based indoor position systems is proposed based on background cloud computing. A novel computing method is applied to the hybrid algorithm for reducing computational complexity. The performance of the hybrid algorithm is simulated with both the example of Nearest Neighbors (NN) and K-Nearest Neighbors (KNN) algorithm, and the computational complexity is analyzed theoretically. Simulation results indicate that the better location performance can be achieved and the computational complexity is reduced by proposed NN-KNN hybrid algorithm, and the NNKNN hybrid algorithm is also suitable for other location systems.