A Fingerprint Positioning Algorithm Based on Multilevel Manual Definition

In recent years, more and more attention has been drawn to the indoor positioning algorithm based on fingerprint model. Although, such algorithm is irrelevant to the location of the signal source, it has the drawbacks of low accuracy and poor real-time capability. This paper proposes a fingerprint positioning algorithm based on multilevel manual definition. In the offline phase of the fingerprint positioning algorithm, source data in the fingerprint database is processed by average weighting method, and a high precision fingerprint database is established to improve matching accuracy. In the online phase, multilevel matching algorithm is employed to shorten matching time. Experiments based on ZigBee technology were conducted, and the results showed that the algorithm for establishing a fingerprint database can improve matching precision by 50% and shorten matching time by 50% compared to the mode of RSSI.