In typical Wi-Fi based indoor positioning systems employing fingerprint model,plentiful fingerprints need to be trained by trained experts or technician,which extends labor costs and restricts their promotion.In this ...In typical Wi-Fi based indoor positioning systems employing fingerprint model,plentiful fingerprints need to be trained by trained experts or technician,which extends labor costs and restricts their promotion.In this paper,a novel approach based on crowd paths to solve this problem is presented,which collects and constructs automatically fingerprints database for anonymous buildings through common crowd customers.However,the accuracy degradation problem may be introduced as crowd customers are not professional trained and equipped.Therefore,we define two concepts:fixed landmark and hint landmark,to rectify the fingerprint database in the practical system,in which common corridor crossing points serve as fixed landmark and cross point among different crowd paths serve as hint landmark.Machinelearning techniques are utilized for short range approximation around fixed landmarks and fuzzy logic decision technology is applied for searching hint landmarks in crowd traces space.Besides,the particle filter algorithm is also introduced to smooth the sample points in crowd paths.We implemented the approach on off-the-shelf smartphones and evaluate the performance.Experimental results indicate that the approach can availably construct WiFi fingerprint database without reduce the localization accuracy.展开更多
A Wi-Fi fingerprinting localization approach has attracted increasing attention in recent years due to the ubiquity of Access Point( AP). However,typical fingerprinting localization methods fail to resist accidental e...A Wi-Fi fingerprinting localization approach has attracted increasing attention in recent years due to the ubiquity of Access Point( AP). However,typical fingerprinting localization methods fail to resist accidental environmental changes,such as AP movement. In order to address this problem,a robust fingerprinting indoor localization method is initiated. In the offline phase,three attributes of Received Signal Strength Indication( RSSI) —average,standard deviation and AP's response rate—are computed to prepare for the subsequent computation. In this way,the underlying location-relevant information can be captured comprehensively. Then in the online phase, a three-step voting scheme-based decision mechanism is demonstrated, detecting and eliminating the part of AP where the signals measured are severely distorted by AP 's movement. In the following localization step,in order to achieve accuracy and efficiency simultaneously,a novel fingerprinting localization algorithm is applied. Bhattacharyya distance is utilized to measure the RSSI distribution distance,thus realizing the optimization of MAximum Overlapping algorithm( MAO). Finally,experimental results are displayed,which demonstrate the effectiveness of our proposed methods in eliminating outliers and attaining relatively higher localization accuracy.展开更多
Wi-Fi technology has become an important candidate for localization due to its low cost and no need of additional installation.The Wi-Fi fingerprint-based positioning is widely used because of its ready hardware and a...Wi-Fi technology has become an important candidate for localization due to its low cost and no need of additional installation.The Wi-Fi fingerprint-based positioning is widely used because of its ready hardware and acceptable accuracy,especially with the current fingerprint localization algorithms based on Machine Learning(ML)and Deep Learning(DL).However,there exists two challenges.Firstly,the traditional ML methods train a specific classification model for each scene;therefore,it is hard to deploy and manage it on the cloud.Secondly,it is difficult to train an effective multi-classification model by using a small number of fingerprint samples.To solve these two problems,a novel binary classification model based on the samples’differences is proposed in this paper.We divide the raw fingerprint pairs into positive and negative samples based on each pair’s distance.New relative features(e.g.,sort features)are introduced to replace the traditional pair features which use the Media Access Control(MAC)address and Received Signal Strength(RSS).Finally,the boosting algorithm is used to train the classification model.The UJIndoorLoc dataset including the data from three different buildings is used to evaluate our proposed method.The preliminary results show that the floor success detection rate of the proposed method can reach 99.54%(eXtreme Gradient Boosting,XGBoost)and 99.22%(Gradient Boosting Decision Tree,GBDT),and the positioning error can reach 3.460 m(XGBoost)and 4.022 m(GBDT).Another important advantage of the proposed algorithm is that the model trained by one building’s data can be well applied to another building,which shows strong generalizable ability.展开更多
基金partially sponsored by National Key Project of China (No.2012ZX03001013-003)
文摘In typical Wi-Fi based indoor positioning systems employing fingerprint model,plentiful fingerprints need to be trained by trained experts or technician,which extends labor costs and restricts their promotion.In this paper,a novel approach based on crowd paths to solve this problem is presented,which collects and constructs automatically fingerprints database for anonymous buildings through common crowd customers.However,the accuracy degradation problem may be introduced as crowd customers are not professional trained and equipped.Therefore,we define two concepts:fixed landmark and hint landmark,to rectify the fingerprint database in the practical system,in which common corridor crossing points serve as fixed landmark and cross point among different crowd paths serve as hint landmark.Machinelearning techniques are utilized for short range approximation around fixed landmarks and fuzzy logic decision technology is applied for searching hint landmarks in crowd traces space.Besides,the particle filter algorithm is also introduced to smooth the sample points in crowd paths.We implemented the approach on off-the-shelf smartphones and evaluate the performance.Experimental results indicate that the approach can availably construct WiFi fingerprint database without reduce the localization accuracy.
基金Sponsored by the National High Technology Research and Development Program of China(Grant No.2014AA123103)
文摘A Wi-Fi fingerprinting localization approach has attracted increasing attention in recent years due to the ubiquity of Access Point( AP). However,typical fingerprinting localization methods fail to resist accidental environmental changes,such as AP movement. In order to address this problem,a robust fingerprinting indoor localization method is initiated. In the offline phase,three attributes of Received Signal Strength Indication( RSSI) —average,standard deviation and AP's response rate—are computed to prepare for the subsequent computation. In this way,the underlying location-relevant information can be captured comprehensively. Then in the online phase, a three-step voting scheme-based decision mechanism is demonstrated, detecting and eliminating the part of AP where the signals measured are severely distorted by AP 's movement. In the following localization step,in order to achieve accuracy and efficiency simultaneously,a novel fingerprinting localization algorithm is applied. Bhattacharyya distance is utilized to measure the RSSI distribution distance,thus realizing the optimization of MAximum Overlapping algorithm( MAO). Finally,experimental results are displayed,which demonstrate the effectiveness of our proposed methods in eliminating outliers and attaining relatively higher localization accuracy.
文摘Wi-Fi technology has become an important candidate for localization due to its low cost and no need of additional installation.The Wi-Fi fingerprint-based positioning is widely used because of its ready hardware and acceptable accuracy,especially with the current fingerprint localization algorithms based on Machine Learning(ML)and Deep Learning(DL).However,there exists two challenges.Firstly,the traditional ML methods train a specific classification model for each scene;therefore,it is hard to deploy and manage it on the cloud.Secondly,it is difficult to train an effective multi-classification model by using a small number of fingerprint samples.To solve these two problems,a novel binary classification model based on the samples’differences is proposed in this paper.We divide the raw fingerprint pairs into positive and negative samples based on each pair’s distance.New relative features(e.g.,sort features)are introduced to replace the traditional pair features which use the Media Access Control(MAC)address and Received Signal Strength(RSS).Finally,the boosting algorithm is used to train the classification model.The UJIndoorLoc dataset including the data from three different buildings is used to evaluate our proposed method.The preliminary results show that the floor success detection rate of the proposed method can reach 99.54%(eXtreme Gradient Boosting,XGBoost)and 99.22%(Gradient Boosting Decision Tree,GBDT),and the positioning error can reach 3.460 m(XGBoost)and 4.022 m(GBDT).Another important advantage of the proposed algorithm is that the model trained by one building’s data can be well applied to another building,which shows strong generalizable ability.
