ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information
<p dir="ltr">Recent transportation mode detection systems propose leveraging signals from only the serving cell tower to ensure ubiquity and practical deployability across all phones. However, existing solutions employ limited statistical hand-engineered features and traditional mach...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | , |
| Published: |
2024
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1864513522144116736 |
|---|---|
| author | Sherif Mostafa (13548916) |
| author2 | Moustafa Youssef (8089592) Khaled A. Harras (23275033) |
| author2_role | author author |
| author_facet | Sherif Mostafa (13548916) Moustafa Youssef (8089592) Khaled A. Harras (23275033) |
| author_role | author |
| dc.creator.none.fl_str_mv | Sherif Mostafa (13548916) Moustafa Youssef (8089592) Khaled A. Harras (23275033) |
| dc.date.none.fl_str_mv | 2024-11-22T15:00:00Z |
| dc.identifier.none.fl_str_mv | 10.1145/3678717.3691250 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/conference_contribution/ModeSense_Ubiquitous_and_Accurate_Transportation_Mode_Detection_using_Serving_Cell_Tower_Information/31443979 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Information and computing sciences Data management and data science Human-centred computing Machine learning Transportation mode estimation Mobility classification Humanactivity recognition Deep learning Feature engineering |
| dc.title.none.fl_str_mv | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| dc.type.none.fl_str_mv | Text Conference contribution info:eu-repo/semantics/publishedVersion text conference object |
| description | <p dir="ltr">Recent transportation mode detection systems propose leveraging signals from only the serving cell tower to ensure ubiquity and practical deployability across all phones. However, existing solutions employ limited statistical hand-engineered features and traditional machine learning classifiers, leading to low estimation accuracy.</p><p dir="ltr">We present ModeSense, a novel ubiquitous and easily deployable transportation mode detection system that achieves high accuracy using handover and received signal strength information from only the serving cell tower. ModeSense introduces a hybrid deep learning-based model that fuses hand-engineered features with automatically extracted features from the serving tower data. This provides ModeSense with a comprehensive set of features, allowing it to overcome the limited information available when using the serving tower only. ModeSense effectively addresses several design challenges, including cellular data issues, absence of motion information, loss of temporal context in long-term dependencies, and optimizing the fusion of features. We extensively evaluate ModeSense using a real-world dataset comprising 395 hours of data. Our results indicate that ModeSense improves average precision and recall by up to 23% and 22%, respectively, compared to state-of-the-art systems. Moreover, ModeSense achieves these enhancements while remaining robust to phone placement variations and user heterogeneity, and requiring over five times less power consumption than GPS.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: Proceedings of the 32nd ACM International Conference on Advances in Geographic Information Systems<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1145/3678717.3691250" target="_blank">https://dx.doi.org/10.1145/3678717.3691250</a></p> |
| eu_rights_str_mv | openAccess |
| id | Manara2_4904b1b1b467691ec6d75cb5b0edb863 |
| identifier_str_mv | 10.1145/3678717.3691250 |
| network_acronym_str | Manara2 |
| network_name_str | Manara2 |
| oai_identifier_str | oai:figshare.com:article/31443979 |
| publishDate | 2024 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower InformationSherif Mostafa (13548916)Moustafa Youssef (8089592)Khaled A. Harras (23275033)Information and computing sciencesData management and data scienceHuman-centred computingMachine learningTransportation mode estimationMobility classificationHumanactivity recognitionDeep learningFeature engineering<p dir="ltr">Recent transportation mode detection systems propose leveraging signals from only the serving cell tower to ensure ubiquity and practical deployability across all phones. However, existing solutions employ limited statistical hand-engineered features and traditional machine learning classifiers, leading to low estimation accuracy.</p><p dir="ltr">We present ModeSense, a novel ubiquitous and easily deployable transportation mode detection system that achieves high accuracy using handover and received signal strength information from only the serving cell tower. ModeSense introduces a hybrid deep learning-based model that fuses hand-engineered features with automatically extracted features from the serving tower data. This provides ModeSense with a comprehensive set of features, allowing it to overcome the limited information available when using the serving tower only. ModeSense effectively addresses several design challenges, including cellular data issues, absence of motion information, loss of temporal context in long-term dependencies, and optimizing the fusion of features. We extensively evaluate ModeSense using a real-world dataset comprising 395 hours of data. Our results indicate that ModeSense improves average precision and recall by up to 23% and 22%, respectively, compared to state-of-the-art systems. Moreover, ModeSense achieves these enhancements while remaining robust to phone placement variations and user heterogeneity, and requiring over five times less power consumption than GPS.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: Proceedings of the 32nd ACM International Conference on Advances in Geographic Information Systems<br>License: <a href="https://creativecommons.org/licenses/by/4.0/" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1145/3678717.3691250" target="_blank">https://dx.doi.org/10.1145/3678717.3691250</a></p>2024-11-22T15:00:00ZTextConference contributioninfo:eu-repo/semantics/publishedVersiontextconference object10.1145/3678717.3691250https://figshare.com/articles/conference_contribution/ModeSense_Ubiquitous_and_Accurate_Transportation_Mode_Detection_using_Serving_Cell_Tower_Information/31443979CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/314439792024-11-22T15:00:00Z |
| spellingShingle | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information Sherif Mostafa (13548916) Information and computing sciences Data management and data science Human-centred computing Machine learning Transportation mode estimation Mobility classification Humanactivity recognition Deep learning Feature engineering |
| status_str | publishedVersion |
| title | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| title_full | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| title_fullStr | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| title_full_unstemmed | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| title_short | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| title_sort | ModeSense: Ubiquitous and Accurate Transportation Mode Detection using Serving Cell Tower Information |
| topic | Information and computing sciences Data management and data science Human-centred computing Machine learning Transportation mode estimation Mobility classification Humanactivity recognition Deep learning Feature engineering |