Improve Accuracy of Fingerprinting Localization with Temporal Correlation of the RSS

Tian, Xiaohua; Wang, Mei; Li, Wenxin; Jiang, Binyao; Xu, Dongyan; Wang, Xinbing; Xu, Jun

doi:10.1109/tmc.2017.2703892

Cited by 35 publications

(15 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is comparable to various attempts reported in literature to improve the fingerprinting accuracy. For instance, the authors of [48] have theoretically demonstrated that the temporal correlation of the RSSI could be exploited to further improve fingerprinting accuracy by 13%-30%. One of the advantages of the proposed method is its applicability when heterogeneous clients are used without the need for time-consuming exhaustive measurements using each device to populate the radiomap.…”

Section: Discussionmentioning

confidence: 99%

Multidevice Map-Constrained Fingerprint-Based Indoor Positioning Using 3-D Ray Tracing

Raspopoulos

2018

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

Abstract-This paper studies the use of deterministic channel modelling through 3D Ray Tracing (RT) for constructing deviceindependent radiomaps for Wi-Fi RSSI-based fingerprinting indoor positioning, applicable to different devices. Device heterogeneity constitutes a limitation in fingerprint-based approaches and also constructing radiomaps through extensive in-situ measurement campaigns is laborious and time-consuming even with a single device let alone the need for radiomaps constructed using multiple different devices. This work tackles both challenges through the use of 3D RT for radiomap generation in conjunction with data calibration using a small set of device-specific measurements to make the radiomap device-independent. The efficiency of this approach is evaluated using simulations and measurements in terms of the time spent to generate the radiomap, the amount of device-specific data required for calibration and in terms of the achievable positioning accuracy. Potential accuracy improvements in the RT-based indoor positioning processes are further investigated, by studying the use of map constraints into the algorithm in the form of a-priori probabilities. In this approach, a Route Probability Factor (RPF), which reflects the likelihood of a user being in various locations inside the environment is used. The outcome of the evaluation process which includes a study of different RPF distributions, indicates the validity of the approach, demonstrated by a reduction in the positioning error for various devices. The versatility of this approach is also demonstrated for different scenarios, different devices and by considering different device-handling conditions.

show abstract

Section: Discussionmentioning

confidence: 99%

Multidevice Map-Constrained Fingerprint-Based Indoor Positioning Using 3-D Ray Tracing

Raspopoulos

2018

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

show abstract

“…Using the linear fitting of discrete fingerprint samples of the m -th T-AP and N APs, the correlation function of m -th T-AP and N APs is constructed [ 5 ], which is recorded as:

where

. Using the RSS data in

and

as measured data, the correlation coefficient

in Equation (7) can be solved by least square method because of

.…”

Section: Underground Fingerprint Positioning System Based On Ma-vamentioning

confidence: 99%

“…Many important works in fingerprint positioning are mission-like noise reduction of fingerprint samples [ 5 ], the optimal selection of access points (APs) [ 6 ], and improvement of fingerprint positioning method [ 7 , 8 ]. However, there is little research on how to further improve the fingerprint positioning accuracy when APs are sparse deployed in coal mine tunnels.…”

Section: Introductionmentioning

confidence: 99%

Underground Coal Mine Fingerprint Positioning Based on the MA-VAP Method

Song

Qian

2020

Sensors

View full text Add to dashboard Cite

The access points (APs) in a coal mine wireless local area network (WLAN) are generally sparsely distributed. It can, with difficulty, satisfy the basic requirements of the fingerprint positioning based on Wi-Fi. Currently, the effectiveness of positioning is ensured by deploying more APs in an underground tunnel, which significantly increases system cost. This problem can be solved by using the Virtual Access Point (VAP) method that introduces virtual access points (VAPs), which can be virtually arranged in any part of the positioning area without installing actual access points. The drawback of the VAP method is that the generated received signal strength (RSS) value of a VAP is calculated based on the mapping of RSS value from only one corresponding access point (AP). This drawback does not consider the correlation between different AP signals and the generated RSS value of a VAP, which makes the modeling of fingerprint samples and real-time RSS collection incomplete. This study proposed a Multi-Association Virtual Access Point (MA-VAP) method takes into account the influence of multi-association. The multi-association coefficient is calculated based on the correlation between the RSS values of a VAP and multiple access points (APs). Then, the RSS value generated by a VAP is calculated using the multi-association function. The real-time collected RSS values from multiple APs related to this VAP are the input of the multi-association function. The influence of the number of VAPs and their arrangement on positioning accuracy is also analyzed. The experimental positioning results show that the proposed MA-VAP method achieves better positioning performance than the VAP method for the same VAP arrangement. Combined with the Weight K-Nearest Neighbors (WKNN) algorithm and Kernel Principal Component Analysis (KPCA) algorithm, the positioning error of the MA-VAP method of the error distance cumulative distribution function (CDF) at 90% is 4.5 m (with WKNN) and 3.5 m (with KPCA) in the environment with non-line-of-sight (NLOS) interference, and the positioning accuracy is improved by 10% (with WKNN) and 22.2% (with KPCA) compared with the VAP method. The MA-VAP method not only effectively solves the fingerprint positioning problem when APs are sparse deployed, but also improves the positioning accuracy.

show abstract

“…Recent years have witnessed a growing interest in exploiting wireless signals for tasks like gait identification [27,76,90,92], gesture recognition [8, 26, 32, 40, 42, 53-57, 62, 66, 71, 93], localization [11,63,64,69,84,88], health and risk assessment [4,7,25,36,38,45,47,48,68], activity detection [10,14,58,59,74,75,77,80,81,89], human detection [18,82,97,98] and emotion recognition [94]. Indeed, wireless sensing has moved from an a research niche [78] to a mainstream activity.…”

Section: Related Workmentioning

confidence: 99%

CrossSense

Zhang

Tang

et al. 2018

Proceedings of the 24th Annual International Conference on Mobile Computing and Networking

188

View full text Add to dashboard Cite

We present CrossSense, a novel system for scaling up WiFi sensing to new environments and larger problems. To reduce the cost of sensing model training data collection, CrossSense employs machine learning to train, off-line, a roaming model that generates from one set of measurements synthetic training samples for each target environment. To scale up to a larger problem size, CrossSense adopts a mixture-of-experts approach where multiple specialized sensing models, or experts, are used to capture the mapping from diverse WiFi inputs to the desired outputs. The experts are trained offline and at runtime the appropriate expert for a given input is automatically chosen. We evaluate CrossSense by applying it to two representative WiFi sensing applications, gait identification and gesture recognition, in controlled single-link environments. We show that CrossSense boosts the accuracy of state-of-the-art WiFi sensing techniques from 20% to over 80% and 90% for gait identification and gesture recognition respectively, delivering consistently good performance-particularly when the problem size is significantly greater than that current approaches can effectively handle. CCS CONCEPTS • Networks → Wireless access networks; • Human-centered computing → Ubiquitous and mobile computing systems and tools;

show abstract

Improve Accuracy of Fingerprinting Localization with Temporal Correlation of the RSS

Cited by 35 publications

References 39 publications

Multidevice Map-Constrained Fingerprint-Based Indoor Positioning Using 3-D Ray Tracing

Multidevice Map-Constrained Fingerprint-Based Indoor Positioning Using 3-D Ray Tracing

Underground Coal Mine Fingerprint Positioning Based on the MA-VAP Method

CrossSense

Contact Info

Product

Resources

About