Research On Indoor Fingerprint Location Based On Spatio-Temporal Constraints

The selection of access point(AP),the construction and update of offline fingerprint database and the online matching operation are seriously affected by the signal refraction,reflection and multipath effect caused by obstacles in indoor environment.The paper uses the temporal and spatial characteristics of the measured data to build an offline fingerprint database based on the spatial distribution of the signal.In order to reduce the maintenance cost of the fingerprint database and improve the online positioning accuracy,it constrains the target motion process by using the temporal characteristics of the observed data.The specific research work of the paper is mainly as follows:(1)For the signal redundancy of similar AP points and the overlapping of coverage areas in the non-autonomous deployment environment,a new AP point selection method based on correlation coefficient and improved information entropy is proposed.According to the concept of correlation coefficient,AP sets with high signal distribution similarity are fused to reduce the overfitting of position estimation caused by AP redundancy.Based on the traditional entropy increase criterion,the proposed method strengthens the influence of signal coverage overlap in the process of AP point selection by adding the coverage redundancy factor,so as to obtain AP sets with relatively scattered coverage area and stable signal distribution.Experimental results show that the proposed algorithm can achieve the signal coverage of the target area under the low dimensional AP set,and maintain excellent positioning accuracy compared with other methods.(2)For the fluctuation of single sampling measurement value and the mutual interference between signals in indoor environment,the paper proposes an indoor positioning system based on the subarea multivariate Gaussian mixture model.According to the AP position and indoor spatial structure,the system uses SVM classification in “one-against-all” form to partition the target area in order to predict the subarea with signal changes.A multivariate Gaussian mixture model based on the mutual interference between signals is established by using the coupling relationship between multiple communication devices in the subarea.It is important to improve the positioning accuracy which affected by signal fluctuation.When the indoor environment changes,the adaptive updating algorithm based on the subarea multivariate Gaussian mixture model can judge the reliability of fingerprint data in each segmentation.Moreover,it can update the model parameters in the subarea with large signal fluctuation by the adaptive algorithm to strengthen the coupling relationship between the model and the existing environment.Experimental result demonstrates that the proposed algorithm can build a stable and maintainable indoor signal distribution model by using a relatively small number of sample data.Its positioning accuracy is also improved to a certain extent compared to other algorithms.(3)For improving the robustness of position system and reducing the location error,the paper proposes a fingerprint positioning method based on recursive Bayesian.To solve the blindness and unreliability of the location fingerprint data in offline phase,the fingerprint database based the sample variance is developed to measure the confidence of sampling values and reduce the impact of environmental factors,which improves the reliability for online localization.The proposed method provides the target position at current time utilizing the Markov model that is established by the constraint relationship between moments in the source movement,which avoids the jump problem of the position estimation and poor robustness and improves the localizaton accuracy.The extensive experimental results demonstrate that the average localizaton error norm of the proposed algorithm achieves significantly lower errors than other traditional schemes.