2019 Springer
2019 Springer
Abstract.Indoor location and tracking systems based on Wi-Fi signals are both gainful and precise, they
are able to attain certain positioning level using existing Wi-Fi infrastructure environment.
The implementation within a wireless sensors network proffer an updating card that estimate true values of
RSSI on the fingerprints points, and combine algorithms localization that offer accurate positions
estimations.
In this paper a localization algorithm will be described it offers a good accuracy using a particular filter
which tracks multiple points to characterize a trajectory. Finally an evaluation of methods performance is
desirable to verify that the system is more accurate and efficient on tracking.
Keywords: Wireless sensor network, Localization algorithm, Indoor location, access point, signal
strength, location tracking
1 Introduction
Indoor localization [1] is one of the most important problems in intelligent services. They are
considered much more difficult than outdoor localization problems because GPS signal is not available
within building or nearby huge structure.
The technologies of indoor location tracking are used in public building [2] like museums, transit
stations or hospitals. In contrast, location errors may result in undesirable deliveries of wrong information
to the wrong people at wrong place.
Location tracking and positioning systems can be classified by: Measurement attribute [3],
localization Algorithms [4] and communication protocol. The classifications by the measurement
techniques are employed to determine mobile device location (localization). Location tracking approach
differs in terms of the specific technique used to precisely detect and measure the position of the mobile
device in the target environment under observation. The Real Time Location Systems (RTLS) as a type
of measurement attribute includes many basic systems able to determine positions using the following
parameters:
• Cell of origin (nearest cell)
• Distance (lateration)
• Angle (angulations)
• Location patterning (pattern recognition)
RTLS solution typically includes location sensors that are attached to various assets, using a unique
ID. The system can locate the tags and report real time information about its positioning within the
facility estimation, based RTLS solution enables tracking unit. This location services are preferment in
finding and tracking assets (objects) and people; they deliver information for users mainly on client
software.
Manufacturers of RTLS systems have combined the functionality of both systems and incorporated
both technologies GPS and Wi-Fi in a device to provide the location data.
System as such utilizes the availability of GPS for outdoor location and switch to Wi-Fi when
confronting obstacles.
Wi-Fi [6] based system potentially works indoor. For instance, Radar is an indoor location system
based on Wi-Fi signal [7], [8]. It uses tow methods to calculate locations from signal information. The
In this paper, the location and the tracking use the Wi-Fi strength signal. It is implemented in a
wireless sensors network to estimate the positions using the fingerprinting location method. The Wi-Fi
locations use the LF and sensors in two aspects: the training one, in charge of collecting different RSS
values due to sensors and their databases. The second aspect is interested in position estimation using
KNN Algorithm [10]. Then we employed the particular filter to track multiple points and characterize a
trajectory. This solution shows many benefits as the system resulted with 3 meter of location precision
adapted with the indoor space.
This paper is organized as follows, in section 2 we describes the design of the proposed Indoor
location system based on wireless sensors networks, next presents location mythology was described. In
section 4, the improvement of trajectory accuracy is due to a particular filter, section 5 gives a
performance evaluation of location estimation. Finally the last section presents discussion of future
works.
The experiment was performed in the hall of the National School of Applied Sciences in Tangier; the
dimension is approximately 25 x 15 m².
We use 6 AP operating on the radio channel of IEEE 802,11b at 2, 54 GHz with 3 overlapping channel.
Also 300 distributed sensors as shown in Figure1.
The received signal strength is in the range of -40dBm and -90dBm, and the highly RSS value is
-25dBm approximately at 1m (APs distance).
Access points are located in the following positions (3; 3), (17; 3), (3; 13), (17; 13), (14; 10), (8; 6).
The location and the tracking system use the Wi-Fi strength signal. It is implement the wireless
sensors network to estimate the positions using the fingerprinting location method.
The Wi-Fi locations algorithm use tree aspects: the training one, in charge of collecting different RSS
values due to sensors and their databases. The second aspect is interested in position estimation using
KNN Algorithm ,finaly In order to improve the accuracy the Particular Filter is implemented
Fig.1. A figure caption is always placed below the illustration. Short captions are centered, while long ones are
justified. The macro button chooses the correct format automatically.
This section describes the algorithm used in the proposed approach: signal propagation theorem, K-
nearest neighbor and a probabilistic estimation.
K-nearest neighbor is applied to specify the sensors positions after collecting the RSSi vector F =
[F1….FN] values.
With N as the number of the deployed AP in the area and Pi as an independent RSSI from a data AP.
In order to find the nearest neighbor we use KNN algorithm, which is a set of strength signal, then
we also use the weighted average of geometric coordinate relative to the K nearest sensors.
𝑘
1 𝑃𝑖 . (𝑥𝑖 , 𝑦𝑖 )
.∑ 𝑘
𝑘 ∑𝑖=1 𝑃𝑖
𝑖=1
Wi-Fi RSS is retrieved from all grid points, the radio frequency signal obeys propagation based
model.
The Wi-Fi RSS research support taking on Gaussian distribution when the size samples RSS is large.
Supposing the case, the variance of Wi-Fi RSS distribution are σs& µs and the probability of getting the
exact location is donate.
In order to improve the accuracy the Particular Filter is implemented, so that we can estimate the
state of a discrete time tracking process monitored by no linear stochastic difference equation. Then the
covariance matrix is used to estimate system evolution that offers more precision than the regular
tracking, which provides only the estimated location correction.
5 Performance evaluation
This part of the work is interested in system performance, and starts by examination of errors
through the mesh, then the effect of varying the number of neighbors, number of access points and also
the number of sensors used, finally we evaluate the estimation of particular filter trajectory.
As shown in the figure above (figure 2), the mesh is fully swept, where the error is calculated in
meters from sensors. This figure shows also a maximal error approximately about 2.7 m and average
error approximately about 1m. So we notice that errors increase in minimum power density positions,
and the received signal from various access points is a source of several errors.
In the study shown in Figure 3, we varied the size of the mesh, also increased the distance
between nodes up to 4.5m. We notice that the more the mesh size increases, the more the error
increases. And the average error attains 1m while the distance between nodes is 1m.
Figure 4 indicates the effect of variation neighbor numbers and shows that the
location becomes more accurate. Referring to the figure, the error is minimized to 0.3m using 8
neighbors.
The nearest neighbor algorithm [10] uses a Map for keeping track of the status of the users in a
particular location. There is a calibration process in which the tracks of all users are kept so that the
location is achieved easily.
The K-Nearest Neighbor (K-NN) algorithm is used to estimate the mobile sensor location. Once the
location is figured, the Particular filter is attributed to improve the performance. Those experiments
examine the accurate trajectory when a user walks around a room with a mobile sensor.
The figure below (Figure5) offers the trajectory estimation using a particular filter algorithm
combined with position already described. We observe that error decreases along the trajectory to an
average of 1m.
In this paper, different techniques stated for location based tracking and also concentrating on a mass
usage of wireless Wi-Fi signal for location tracing. Wi-Fi signals allowing an immense field of services
to track the user’s location. In this scenario the distance between sensors is a major issue.
We proposed and describe a localization algorithm in this research paper that offers a good accuracy
and stability using a particular filter. Many improvements could be made such as reducing the number
of sensors using clustering; also using other metrics give more precision than Euclidian distance.
References
1- Hyo-Sung Ahn, and Wonpil Yu2, Indoor Localization Techniques based on Wireless Sensor
Networks.korea
2- Farid, Rosdiadee Nordin, andMahamod Ismail, Recent Advances in Wireless Indoor Localization
Techniques and System.
3- Klaithem Al Nuaim, Hesham Kamel, A Survey of Indoor Positioning Systems and Algorithms, 2011
International Conference on Innovations in Information Technology
4- Han Zou , Xiaoxuan Lu, Hao Jiang and Lihua Xie, A Fast and Precise Indoor Localization Algorithm
Based on an Online Sequential Extreme Learning Machine, Sensors 2015.
5- Jun-Ho Huh and Kyungryong Seo, An Indoor Location-Based Control System Using Bluetooth Beacons
for IoT Systems.
6- Atul Gosai. Ph.D, Rushi Raval, Real Time Location based Tracking using WIFI Signals, International
Journal of Computer Applications (0975 – 8887) Volume 101– No.5, September 2014.
7- An Overview Shixiong Xia , Yi Liu , Guan Yuan , Mingjun Zhu and Zhaohui Wang. Indoor Fingerprint
Positioning Based on Wi-Fi.
8- N. Suguna1 , and Dr. K. Thanushkodi, An Improved k-Nearest Neighbor Classification Using Genetic
Algorithm. IJCSI International Journal of Computer Science Issues, Vol. 7, Issue 4, No 2, July 2010.
9- Sinno Jialin Pan, VincentWenchen Zheng, Qiang Yang and Derek Hao Hu, Transfer Learning forWiFi-
based Indoor Localization, 2008, Association for the Advancement of Artificial Intelligence.
10- Lu-Chuan Kung Jennifer C. HouY, Eddie C.L. Chan, George Baciu, S.C. Mak, Using Wi-Fi Signal
Strength to Localize in Wireless Sensor Networks, 2009 International Conference on Communications
and Mobile Computing, Zero-configuration indoor localization over IEEE 802.11 wireless Infrastructure,
Springer Science+Business Media, LLC 2008.
11- Li-wei Chan, Ji-rung Chiang, Yi-chao Chen, Chia-nan Ke, Jane Hsu, Hao-hua Chu, Collaborative
Localization: Enhancing WiFi-Based Position Estimation with Neighborhood Links in Clusters.
12- Wei-Peng Chen, Jennifer C. Hou and Lui Sha, “Dynamic Clustering for Acoustic Target Tracking in
Wireless Sensor Networks”, IEEE TRANSACTIONS ON MOBILE COMPUTING, IEEE Computer
Society, 2004, pp. 258-271.
13- Aline Baggio and Koen Langendoen. Monte-carlo localization for mobile wireless sensor networks, In
MSN06, 2006, pp. 317-328.
14- Junqiu Wang, Hongbin Zha, and Roberto Cipolla, “Coarse to-Fine Vision-Based Localization by
Indexing Scale-Invariant Features”, IEEE TRANSACTIONS ON SYSTEMS, MAN, AND
CYBERNETICS—PART B: CYBERNETICS, 2006, vol. 36, pp. 413
15- Hyo-Sung Ahn and Wonpil Yu, Indoor Localization Techniques based on Wireless Sensor Networks.
16- Yi-Chao Chen, Ji-Rung Chiang, Hao-hua Chu, Polly Huang, and Arvin Wen Tsui, Sensor-Assisted Wi-Fi
Indoor Location System for Adapting to Environmental Dynamics, Department of Computer Science and
Information Engineering, Graduate Institute of Networking and Multimedia, Department of Electrical
Engineering, National Taiwan University, Industrial Technology Research Institute.
17- Jeongkeun Lee, Kideok Cho, Seungjae Lee, Taekyoung Kwon, Yanghee Choi. Distributed and energy-
efficient target localization and tracking in wireless sensor networks. School of Computer Science and
Engineering, Seoul National University, San 56-1 Shilim-dong, Kwanak-gu, Seoul, Republic of Korea.
18- Antti Ropponen, Matti Linnavuo, Raimo Sepponen, LF Indoor Location and Identification System,
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 2, NO.
1, MARCH 2009.
19- Scott Bell, Wook Rak Jung, Vishwa Krishnakumar, WiFi-based Enhanced Positioning Systems:
Accuracy Through Mapping, Calibration, and Classification.
20- Lingxuan Hu David Evans, Localization for Mobile Sensor Networks, Department of Computer Science
University of Virginia, Charlottesville.
21- JULUOS. O Smith, JONATHAN. S Label. Closed-Form Least Squares Source Location Estimation from
Range Difference Measurements, IEEE TRANSACTIONS ON ACOUSTICS, SPEECH, AND SIGNAL
PROCESSING, VOL. ASSP-35, NO. 12, DECEMBER 1987.
22- Y. Gwon, R. Jain, T. Kawahara, Robust Indoor Location Estimation of Stationary and Mobile Users,
DoCoMo Communications Laboratories USA.
23- Yi-Chao Chen, Ji-Rung Chiang, Hao-hua Chu, Polly Huang, Arwin Wen Tsui. Sensor Assisted Wi-Fi
Indoor Location System for Adapting to Environment Dynamics, International Workshop on Modeling