A novel methodology to estimate a measurement of the
inherent difficulty of an indoor localization radio map
Sansano-Sansano, Emilio; Montoliu, Raúl; Torres-Sospedra, Joaquín.
Published in: 2017 8th International Conference on Indoor Positioning and Indoor Navigation (IPIN) [139]
Date of Conference: 18-21 Sept. 2017 doi: 10.1109/IPIN.2017.8115939
The variables used to measure indoor localization methods’ accuracy are dependent on the radio map used to test them. The estimated error for a positioning technique strongly depends on the characteristics of the scenario where it has been tested. Variables such as the dimmensions of the scenario, the number of available beacons or their positions, among many others, may affect the expected error for a given data set. Thus, since the estimated error not only depends on the method performance, but it is strongly related to
the scenario attributes, a definitive conclusion on the method’s performance cannot be obtained. This makes it hard to compare different methods’ results. This work presents a novel methodology to obtain a measure of the inherent difficulty of a scenario to obtain accurate localization results when testing an indoor positioning method. The proposed indicator can be used to obtain a difficulty measure from a fingerprinting data set. This indicator will show if the precision obtained with a positioning method, using that data set, can be considered a reliable measurement of the method’s performance. It can be used to provide a measurement of how difficult it would be to get a good position estimation using a given data set of fingerprints. Therefore, it provides a fairer way to compare the performance of two different algorithms evaluated in different environments, as long as the difficulty of the respective radio maps used to test their accuracy is similar. It can also be used as a metric to measure the impact of any modification of the radio map, as the addition or removal of observations.
A New Methodology for Long-Term Maintenance of WiFi
Fingerprinting Radio Maps
Montoliu, Raúl;Sansano-Sansano, Emilio; Belmonte-Fernández, Óscar;
Torres-Sospedra, Joaquín.
Published in: 2018 9th International Conference on Indoor Positioning and Indoor Navigation (IPIN) [102]
Date of Conference: 24-27 Sept. 2018 doi: 10.1109/IPIN.2018.8533825
Despite the benefits of WiFi fingerprinting techniques, one of the main problems of this technique for indoor positioning systems is the radio map maintenance. It is well known that the creation of the radio map is a tedious and long-time task. Besides, if after its creation, some access points are removed from the environment, the accuracy of the system can be dramati- cally affected. The fingerprint used to locate the user could be composed of values produced by a set of access points different from the one used to create the radio map. A common approach to deal with this situation is to
use just the common access points received at the two different moments in time. This process discards the use of some useful information and therefore, the accuracy of the IPS can be drastically reduced.
This work proposes a new methodology to deal with this problem using regression-based imputation techniques. The main hypothesis is that there is a relationship in the signal strength values obtained for each access point with respect to the other existing access points in the environment. The regression techniques can take advantage of these correlations to impute a valid RSSI value for the removed access point. This paper presents an extensive set of experiments comparing different imputation techniques to demonstrate the benefits of using the proposed approach, showing that it can reduce the localization error in almost one meter with respect to a well-known solution.
Improving Positioning Accuracy in Ambient Assisted Liv-
ing Environments. A Multi-Sensor Approach
Sansano-Sansano, Emilio; Belmonte-Fernández, Óscar; Montoliu, Raúl;
Gascó-Compte, Arturo; Caballer-Miedes, Antonio; Bayarri-Iturralde, Pilar Published in: 2019 15th International Conference on Intelligent Environments (IE) [138]
Date of Conference: 24-27 June 2019 doi: 10.1109/IE.2019.00004
The primary purpose of this research is to examine the viability of leveraging other sensors in aiding a WiFi fingerprinting-based positioning system to provide more accurate predictions. In particular, the experiments presented in this paper show that the use of Inertial Motion Units (IMUs), which are present by default in smart devices such as smart-phones or smart-watches, can increase the performance of indoor positioning systems in AAL environ- ments. Furthermore, this paper assesses complementary strategies such as data scaling and the use of consecutive WiFi scanning to further improve the reliability of the indoor positioning systems’ predictions. This research shows that a robust positioning estimation can be derived from such strategies.
Moreover, this can be done without compromising important aspects such as battery duration or unobtrusiveness. This work also explores possible actions to reduce the influence of WiFi signal uncertainty as well as to select the most appropriate machine learning algorithm for the positioning system. The results obtained from the set of experiments presented in this work show an improved accuracy in room detection when using strategies such as data scaling and the use of consecutive WiFi scanning. The results also demonstrate that the use of a significant motion sensor along with the WiFi fingerprints can help to significantly increase the performance of indoor positioning systems.
Evaluation of Crowdsourcing Wi-Fi Radio Map Creation in
a Real Scenario for AAL Applications
Belmonte-Fernández, Óscar; Gascó Compte, Arturo; Sansano-Sansano, Emilio; Quinde, Mario; Giménez Manuel, José Ginés; Augusto Juan Carlos.
Published in: 2019 15th International Conference on Intelligent Environments (IE) [23]
Date of Conference: 24-27 June 2019 doi: 10.1109/IE.2019.00005
Technology can be integrated into the health care of senior citizens to provide safe, high-quality lives, improving their health and happiness, and enabling a longer period of independent living. Indoor positioning technologies are destined to play an important part in these applications.
One of the main drawbacks of WiFi fingerprinting methods is the temporal cost involved in creating a radio map. Crowdsourcing strategies have been presented as a way to minimize the cost of radio map creation. This research presents an extensive study of the issues involved when using crowdsourcing strategies for that purpose. The results provided by extensive experiments performed in a real scenario by three users during two weeks show how join- ing data gathered by different users can improve the accuracy performance of a WiFi-based indoor location.
To prevent issues related to device diversity, the same device model was used by all users. To assess the location accuracy, the KNN, Bayes Network, and Random Forest machine learning algorithms have been tested. The results obtained in this study allow us to conclude that crowdsourcing data improves the accuracy of the location. The results show the feasibility of crowdsourcing data to create radio maps for the indoor location. One the second hand, accuracy decay along time was reported.
Senior Monitoring: A Real Case of Applying a WiFi
Fingerprinting-based Indoor Positioning Method for Peo-
ple Monitoring
Montoliu, Raúl; Sansano-Sansano, Emilio; Gascó-Compte, Arturo; Bel-
monte Fernández, Óscar; Caballer-Miedes, Antonio.
Published as Work-in-Progress paper in: 2019 10th International Conference on Indoor Positioning and Indoor Navigation (IPIN) [100]
Date of Conference: 30 September - 3 October 2019
This work showcases a real example of applying a Wi-Fi fingerprinting-based indoor localization system for monitoring elder people in their homes. The presented system is part of a broad project called Senior Monitoring where the main aim is to monitor elders to study behavioral patterns as a tool for early detection of some cognitive decay diseases. Since the system is used by real users, many situations can not be controlled by system developers and can be a source of errors. This study presents some of the problems arisen when real non-expert users use localization systems, and discuss some strategies to deal with such situations.
The experiments were conducted by 17 volunteers for two months on average in real scenarios, where the conditions are not controlled by the researchers. Participants had to create the radio map of his/her own house following the instructions provided by the system’s developers. The present work contributes to a better understanding of the difficulties and problems that arise when implementing an indoor positioning system in real scenarios with real users.