COhave research

Modeling and simulation of user movements with bio-inspired computing.
Mobile applications for sensing and collection data.

Download GPS/WiFi dataset See our publications

User Experience in GPS Trajectory Search using Pattern

We have proposed, implemented and compared several approaches for user movement (trajectory) extraction. Unlike other approaches, our approaches are purely based on WiFi sensing without the knowledge of user's physical location. This is a favorable approach in scenarios that aim at high energy efficiency. We only collect WiFi information passively, i.e. we only listen to broadcast beacons and do not transmit any probe requests.

Maroš Čavojský, Martin Drozda. User Experience Meets GPS Trajectory Search. IEEE International Conference on Systems, Man, and Cybernetics (SMC) 2023 Maroš Čavojský, Martin Drozda. Search by Pattern in GPS Trajectories. EAI MobiCASE 2022, 13th EAI International Conference on Mobile Computing, Applications and Services. Springer Nature Switzerland, 2022. p. 117-132.

Try search

Leaflet | Map data © OpenStreetMap contributors, Imagery © Mapbox, COhave Dataset

Comparison of GPS Trajectories with the Needleman-Wunsch Algorithm

Our goal was to test if the Needleman-Wunsch algorithm (NWA) for sequence alignment can be efficiently applied to comparing user trajectories, where user locations are provided by Global positioning system (GPS). We proposed several adjustments to this algorithm. Our results show that NWA can be successfully applied to user trajectory comparison.

Maroš Čavojský, Martin Drozda, Zoltán Balogh. Analysis and Experimental Evaluation of the Needleman-Wunsch Algorithm for Trajectory Comparison. Expert Systems with Applications, 2020, Elsevier. Maroš Čavojský, Martin Drozda. Comparison of User Trajectories with the Needleman-Wunsch Algorithm. MobiCASE 2019, 10th EAI International Conference on Mobile Computing, Applications and Services. Springer, Cham. 2019. p. 141-154.

Try comparison


4+ years, 500+ devices,
50 000 000+ GPS, WiFi records

From our experiments we offer data set, that we collected with 455 mobile devices distributed among our students at university. Locations and WiFi networks are collected worldwide.

Year GPSv1 GPSv2 WiFi
2015 6 530 019 - -
2016 13 183 118 460 199 12 287 597
2017 441 786 463 671 13 425 877
2018 83 633 212 971 7 439 015
2019 - 54 403 1 622 007
2020 - 13 145 622 945

Download dataset

You can download report from WiFi dataset analysis written in Slovak language here.

Map tiles by Stamen Design, under CC BY 3.0. Data by OpenStreetMap, under ODbL.

Modeling movement based on WiFi

We have proposed, implemented and compared several approaches for user movement (trajectory) extraction. Unlike other approaches, our approaches are purely based on WiFi sensing without the knowledge of user's physical location. This is a favorable approach in scenarios that aim at high energy efficiency. We only collect WiFi information passively, i.e. we only listen to broadcast beacons and do not transmit any probe requests.

Maroš Čavojský, Marek Uhlar, Marian Ivanis, Martin Molnar, Martin Drozda. User Trajectory Extraction Based on WiFi Scanning. 2018 6th International Conference on Future Internet of Things and Cloud Workshops, pp. 115-120, IEEE, 2018

Try visualisation


Trajectory recording using WiFi scanning

Mobile devices are equipped with a GPS receiver, but its frequent use can cause a high battery drain. Our design of continuous location tracking aims at providing a reasonable trade-off between energy consumption, location acquisition accuracy. We use WiFi scanning for decisions about when to start using GPS receiver. The experiments show that our design provides significant energy savings when compared to other continuous location tracking alternatives. Maroš Čavojský and Martin Drozda. Energy Efficient Trajectory Recording of Mobile Devices Using WiFi Scanning. The First International Workshop on Crowd Intelligence for Smart Cities. Proc. of 2016 Intl. IEEE Conferences on UIC/ATC/ScalCom/CBDCom/IoP/SmartWorld Congress, pp. 1079-1085, IEEE, 2016.

Lite visualisation


WiFi Places Markov Chain (BiGram) Try it


Explore device collected statistics

Explore deep statistics about device collected data. Useful as start for other tools, when you discover dates and times of sensing of device.

Explore statistics