User Aided Self-growing Approach on Radio Map Construction for WLAN Based Localization

D. Zou, W. Meng, S. Han, Z. Gong, B. Yu

Abstract:	Wireless Local Area Networks (WLAN) based finger print positioning technology is one of the most promising indoor positioning solutions [1]. However, it costs too much workload to build the radio map in the offline process, which is considered as the most significant disadvantage of this kind of technique. To overcome this problem, researchers have investigated various methods for fast establishing radio maps [2] [3] [4]. In this paper, a “self-growing algorithm” is proposed to reduce the labor cost in the procedure of the radio map establishment. The proposed scheme is formed as follows. Firstly, Reference Points (RPs) are selected regularly in the service domain. The area near the entrance of the scenario is called “seed domain”. Positioning service is available there as the RPs there are well tested, while the rest RPs are left blank. When user leaves but not far from the “seed domain”, the tracking process [5] ensures the positioning results trustworthy. And the RSS vectors are recorded simultaneously by the Users’ Equipment (UE). The forecasted position combining with the corresponding RSS vector forms one “sample”. The samples near the “seed domain” will be uploaded to the server. The RPs near the “seed domain” can be estimated based on the nearby samples. Furthermore, new samples can enhance the accuracy of the nearby RPs. The estimated RPs with high accuracy is called “grown RPs”, which will be added into the radio map. Thus, the coverage of the “seed domain” would be expanded. More samples would be collected with the increasing of users. By repeating the mentioned process, more “grown RPs” would be filled into the radio map, and the coverage of the radio map would be enlarged. The enlarging process is just like the growing process of plants. Compared to the traditional finger print positioning systems, the proposed system requires an uplink and a server to manage the samples collected by the users. In the proposed scheme, the k-Nearest Neighbor (kNN) algorithm [6] is used for the RP estimation. The spatial distances between the target RP and neighbor samples are calculated, the RSS vectors of the k nearest samples are averaged to estimate the RSS vector of the target RP. The error probability of the kNN algorithm is affected by the positioning bias of the tracking algorithm "e", the "k" value of the kNN algorithm, and the density of the samples "s". The relationship among these parameters is presented in the paper. The quality of the recovered radio map is measured by detecting the similarity with the “target radio map”, which is well established and covered the whole experiment scenario. A performance factor called “Euclidean Distances between Radio Maps (EDRM)” is firstly proposed in this paper. This factor is achieved by calculating the mean value of the Euclidean distances between the “target radio map” RPs and the corresponding RPs in the recovered radio map. The effectiveness of the “growing algorithm” is proved by a radio map establishing simulation. The simulation is taken in a complex indoor scenario established based on the classic indoor path loss model. In the simulation, a large number of samples are placed randomly, and the NN, kNN, and Weighted-kNN (WkNN) algorithms are used separately to recover the radio map. And the simulation results show that the EDRM is proportional to "e". When "k" is greater than 7 or "s" is greater than 40, EDRM approaches its minimum. Since the “target radio map” is absent in the “growing” process, the server can judge the necessity of refreshing the RSS vector of a certain RP by comparing the average spatial distances between the k samples and the RP. In order to reflect a growing process, the growth experiment is carried out in a virtual open hall, and the users’ trajectory is generated randomly. The positioning bias is set according to the results of the former hardware experiment [5]. The experiment results indicate that the proposed method can reduce the workload of the offline process, and the well tested RPs’ amount is decreased from 1500 to 100. The main contributions of this paper are: a User Aided Self-growing Approach on Radio Map Construction for WLAN based Localization is proposed for radio map fast establishment; a performance factor EBRM is first proposed to measure the quality of the recovered radio map. The proposed method is particularly suitable for reducing the workload of the radio map construction in personnel-intensive area, where the commercial positioning service is highly required. The factor EBRM can be used to evaluate the performance of a certain fast establishing method for the radio map. Reference [1] H. Liu, H. Darabi, P. Banerjee, J. Liu, “Survey of Wireless Indoor Positioning Techniques and Systems” IEEE Trans. on Systems, Man, and Cybernetics, vol. 37, no. 6, pp. 1067-1080, November 2007. [2] Y. Kim, Y. Chon, H. Cha, “Smartphone-Based Collaborative and Autonomous Radio Fingerprinting” IEEE Trans. on Systems, Man, and Cybernetics, vol. 42, no. 1, January 2012, pp. 1067-1080. [3] P. M.Scholl, S. Kohlbrecher, V. Sachidananda, K. V. Laerhoven, “Fast Indoor Radio-Map Building for RSSI-based Localization Systems” Networked Sensing Systems (INSS), Antwerp Belgium, June 2012, pp. 1-2. [4] L. Peng, J Han. , W. Meng, J. Liu, “Research on Radio-Map Construction in Indoor WLAN Positioning System” Pervasive Computing Signal Processing and Applications (PCSPA), Harbin China, Sept. 2010, pp. 1073-1077. [5] J. Han, L. Ma, Y. Xu, Z. Deng, “Auxiliary particle filter-based WLAN indoor tracking algorithm” Communications and Networking in China (CHINACOM), Harbin China, Aug. 2011,pp. 600-603. [6] V. HONKAVIRTA, T. PERALA, S. ALI-LOYTTY, R. PICHE, “A Comparative Survey of WLAN Location Fingerprinting Methods” Navigation and Communication 2009 (WPNC’09), Hannover Germany, March 2009, pp. 243-251.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	991 - 997
Cite this article:	Zou, D., Meng, W., Han, S., Gong, Z., Yu, B., "User Aided Self-growing Approach on Radio Map Construction for WLAN Based Localization," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 991-997.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In