Study on fingerprint positioning network based on radio map encoding
1
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Wi-Fi fingerprints are widely applied in indoor positioning as the indoor environment is generally non-line-of-sight (NLOS). To improve their performance through combination with deep learning, further studies are being actively conducted. Exist-ing deep learning-based fingerprints require relearning when the positioning environment changes owing to the radio map, which is a signal strength pattern, being directly memorized using weights. In this paper, we propose a fingerprint positioning network based on radio map encoding that can rebuild the positioning systems with only a small amount of additional learning by solving this data dependency. The proposed network comprises an encoding network that uses a radio map as a feature vector and a location network that estimates the location based on the measured Wi-Fi signal. By vectorizing the radio map, the encoding network can recover by only re-measurement of the service space even if the environment changes. In addition, as the positioning network estimates the cur-rent location by considering the past location, the estimation accuracy is high even when the noise in the Wi-Fi signal becomes se-vere. This study demonstrated the superiority of the proposed network by comparing it with two other networks and improving the maximum error distance of 1.28 m.
Abstract
Wi-Fi fingerprints are widely applied in indoor positioning as the indoor environment is generally non-line-of-sight (NLOS). To improve their performance through combination with deep learning, further studies are being actively conducted. Exist-ing deep learning-based fingerprints require relearning when the positioning environment changes owing to the radio map, which is a signal strength pattern, being directly memorized using weights. In this paper, we propose a fingerprint positioning network based on radio map encoding that can rebuild the positioning systems with only a small amount of additional learning by solving this data dependency. The proposed network comprises an encoding network that uses a radio map as a feature vector and a location network that estimates the location based on the measured Wi-Fi signal. By vectorizing the radio map, the encoding network can recover by only re-measurement of the service space even if the environment changes. In addition, as the positioning network estimates the cur-rent location by considering the past location, the estimation accuracy is high even when the noise in the Wi-Fi signal becomes se-vere. This study demonstrated the superiority of the proposed network by comparing it with two other networks and improving the maximum error distance of 1.28 m.
참고문헌 (31)
- V. Anagnostopoulos, "Gaze-informed loca-tion-based services", International Journal of Geographical Information Science, 31(9), pp.1770~1797 (2017)
- S. Ray, "High performance location-based services in a main-memory database", GeoInformatica, 21(2), pp.293~322 (2017)
- W. Le, "A novel Wi-Fi indoor positioning method based on genetic algorithm and twin support vector regres-sion", The 26th Chinese Control and Decision Conference (2014 CCDC), IEEE, pp.4859~4862 (2014)
- J. Machaj, "Rank based fingerprint-ing algorithm for indoor positioning", 2011 International Conference on Indoor Positioning and Indoor Navigation, IEEE, pp.1~6 (2011)
- A. Montaser, "RFID indoor location iden-tification for construction projects", Automation in Con-struction, 39, pp.167~179 (2014)
- L. Calderoni, "Indoor localization in a hospital environment using random forest classifiers", Expert Systems with Applications, 42(1), pp.125~134 (2015)
- E. García, "DS-UWB indoor positioning system implementation based on FPGAs", Sensors and Actuators A : Physical, 201, pp.172~181 (2013)
- H. Li, "Low-cost 3D bluetooth indoor positioning with least square", Wireless Personal Communications, 78(2), pp.1331~1344 (2014)
- J. V. Marti, "Localization of mobile sensors and actuators for intervention in low-visibility conditions : The ZigBee fingerprinting approach", International Journal of Distributed Sensor Networks, 8(8), pp.951213~ (2012)
- S. G. Nagarajan, "Geo-spatial loca-tion estimation for Internet of Things(IoT)networks with one-way time-of-arrival via stochastic censoring", IEEE Internet of Things Journal, 4(1), pp.205~214 (2016)
- M. Shalaby, "Perfor-mance enhancement of TOA localized wireless sensor networks", Wireless Personal Communications, 95(4), pp.4667~4679 (2017)
- O. Daniel, "Dela-accuracy tradeoff in opportunistic time-of-arrival localization", IEEE Signal Processing Letters, 25(6), pp.763~767 (2018)
- C. Wu, "Automatic radio map adap-tation for indoor localization using smartphones", IEEE Transactions on Mobile Computing, 17(3), pp.517~528 (2017)
- C. Luo, "Pallas : Self-bootstrapping fine-grained passive indoor lo-calization using Wi-Fi monitors", IEEE Transactions on Mobile Computing, 16(2), pp.466~481 (2016)
- J. Zuo, "Multi-phase finger-print map based on interpolation for indoor localization us-ing iBeacons", IEEE Sensors Journal, 18(8), pp.3351~3359 (2018)
- Y. Zhuang, "A two-filter integration of MEMS sensors and Wi-Fi fingerprinting for indoor positioning", IEEE Sensors Journal, 16(13), pp.5125~5126 (2016)
- J. Torres-Sospedra, "Ensembles of indoor position-ing systems based on fingerprinting : Simplifying parame-ter selection and obtaining robust systems", 2016 Interna-tional Conference on Indoor Positioning and Indoor Navi-gation (IPIN), IEEE, pp.1~8 (2016)
- K. Chen, "Slide : Towards fast and accurate mobile fingerprinting for Wi-Fi indoor positioning systems", IEEE Sensors Journal, 18(3), pp.1213~1223 (2017)
- Q. Chen, "Indoor pedestrian track-ing with sparse RSS fingerprints", Tsinghua Science and Technology, 23(1), pp.95~103 (2018)
- J. H. Seong, "Selective unsupervised learn-ing-based Wi-Fi fingerprint system using autoencoder and GAN", IEEE Internet of Things Journal, 7(3), pp.1898~1909 (2019)
- X. Jiang, "FSELM : Fusion semi-supervised extreme learning machine for indoor localization with Wi-Fi and bluetooth fingerprints", Soft Computing, 22(11), pp.3621~3635 (2018)
- X. Wang, "CSI-based fingerprinting for indoor lo-calization : A deep learning approach", IEEE Transactions on Vehicular Technology, 66(1), pp.763~776 (2016)
- R. S. Campos, "Wi-Fi multi-floor indoor positioning considering architectural as-pects and controlled computational complexity", Expert systems with applications, 41(14), pp.6211~6223 (2014)
- J. H. Seong, "Real-time recursive finger-print radio map creation algorithm combining Wi-Fi and geomagnetism", Sensors, 18(10), pp.3390~ (2018)
- H. Li, "Achieving privacy preservation in Wi-Fi fingerprint-based localiza-tion", IEEE INFOCOM 2014-IEEE Conference on Com-puter Communications, IEEE, pp.2337~2345 (2014)
- Z. C. Lipton, "A critical re-view of recurrent neural networks for sequence learning", arXiv preprint, arXiv : 1506. 00019, ~ (2015)
- S. Hochreiter, "Long short-term memory", Neural Computation, 9(8), pp.1735~1780 (1997)
- K. Cho, "Learning phrase representations using RNN encoder-decoder for statistical machine translation", arXiv preprint, arXiv : 1406. 1078, ~ (2014)
- M. Schuster, "Bidirectional recurrent neural networks", IEEE transactions on Signal Processing, 45(11), pp.2673~2681 (1997)
- D. Li, "A feature-scaling-based k-nearest neighbor algorithm for indoor positioning systems", IEEE Internet of Things Journal, 3(4), pp.590~597 (2015)
- M. T. Hoang, "Recurrent neural networks for accurate RSSI indoor localization", IEEE Internet of Things Journal, 6(6), pp.10639~10651 (2019)