Indoor Location Technology And Research Based On Data Mining

With the rapid development of wireless communication technology,the research of indoor location service is also constantly updated.Different from the outdoor environment,the indoor environment is more complex and the space is smaller.At present,the indoor fingerprint location technology based on Wi-Fi has gradually become a research hotspot because of its low deployment cost,flexible networking,easy to realize and other characteristics.Combining with data mining technology for location matching has a good prospect.The main work and innovation of this paper are as follows:First,the KNNDB algorithm based on data mining technology is proposed.Firstly,the existing indoor localization technology and data mining technology are analyzed and compared.KNN is one of the most commonly used matching algorithms in the positioning stage,but KNN algorithm can not remove interference fingerprints,the localization results are not ideal.The KNNDB algorithm proposed in this paper combines KNN algorithm with DBSCAN algorithm,and removes interference fingerprints before localization and matching.The simulation results show that the optimized KNNDB algorithm can get more accurate positioning results.Secondly,a fusion method of CSI and RSSI based on K-means clustering algorithm is proposed.Based on the single received signal strength affected by the external environment and multipath effect,there will be some fluctuations,while CSI is less affected by the external environment.In this paper,the K-means clustering algorithm based on CSI and RSSI fusion location method,combined with RSS and CSI data,and using k-means algorithm for location matching,this method is smaller than the traditional algorithm location error.Thirdly,the O-SVM algorithm is proposed for indoor location,as in the complex indoor environment,the gap between two RSS is not equal to the physical space gap.The O-SVM algorithm proposed in this paper uses the options algorithm to cluster and divide the original data,and assigns different weights to the RSS fingerprints of different regions to get the effective fingerprint database.Then,the SVM matching algorithm is used to locate and match the unknown points with the previous weight values to get the location results.The results show that the optimized algorithm can improve the localization accuracy and is more suitable for localization estimation than the traditional algorithms.