Research And Implementation Of Indoor Wireless Location Technology Based On Embedded System

With the rapid development of wireless communication technology and Internet of things technology and with the continuous popularization of intelligent terminals.people demand for location information is growing constantly, and the wireless location technology is the key technology to achieve the location information services, although the location of the demand for the outdoor environment has been met. However, due to the complex indoor environment, the wireless signal is affected by the reflection, refraction, diffraction and multipath environment,the indoor positioning effect is unsatisfactory.In addition,the use of its cost is high,poor the positioning accuracy of it and other reasons limits the popularization and development of indoor location services. The current research mainly focuses on the theory and technology based on fingerprint positioning, but the positioning accuracy, stability and other aspects are difficult to make people satisfied.This paper first introduces the research background and significance of wireless location technology from the point of view of wireless location technology, diversity and practicability, indoor positioning technology research direction and theory summarized the present research situation at home and abroad,then the related theory of indoor wireless location knowledge were summarized,and the positioning algorithm based on fingerprint is studied.From the point of view of practical application,on the basis of analyzing and researching on the the wireless signal indoor propagation loss model,we analysis the impact of wall loss of signal propagation model.Compensation method is proposed to partition the nonlinear loss, determine the relationship between RSS and transmission distance.And according to the characteristics of indoor wireless signal propagation characteristics and RBF neural network to approximate nonlinear and global optimization, the indoor location algorithm based on radial basis function neural network algorithm,and then the fusion weighted centroid algorithm is used for position estimation.So the problem of the nonlinear localization performance is solved well. The experimental results show that the algorithm can effectively restrain the error caused by indoor multipath propagation environment, the performance of the algorithm is greatly improved compared with the traditional algorithm.In the study of location method based on fingerprint database, support vector machine is introduced, based on the support vector machine (SVM) has the characteristics of good learning and generalization ability.the location is devised into a multi-class classification problem.The support vector machine classification one sidedness of reverse K nearest neighbors method is introduced, the location of the fingerprint classification, fusing classification method of support vector machine classification and reverse K nearest neighbor classification hyperplane, near to SVM by reverse K nearest neighbor points are re classified, reduces the SVM classification hyper plane near the classification error rate, can effectively improve the performance of classifier, the classification accuracy has been improved. The test results show that the positioning accuracy of the positioning algorithm, the result has higher volatility, and effectively reduces the location results.Finally,we designed an simple indoor positioning system based on embedded system according to the fingerprint location algorithm in this paper, and tested the system performance and summarized the results.