Cláudio Diego T. de Souza , José Ferreira de Rezende , Carlos Alberto V. Campos
{"title":"Federated Learning assisted framework to periodically identify user communities in urban space","authors":"Cláudio Diego T. de Souza , José Ferreira de Rezende , Carlos Alberto V. Campos","doi":"10.1016/j.adhoc.2024.103589","DOIUrl":null,"url":null,"abstract":"<div><p>Identifying individuals with similar behaviors and mobility patterns has become essential to improving the functioning of urban services. However, since these patterns can vary over time, such identification needs to be done periodically. Furthermore, once mobility data expresses the routine of individuals, privacy must be guaranteed. In this work, we propose a framework for periodically detecting and grouping individuals with behavioral similarities into communities. To accomplish this, we built an autoencoder model to extract spatio-temporal mobility features from raw user data at periodic intervals. We used Federated Learning (FL) as a training approach to preserve privacy and alleviate time-consuming training and communication costs. To determine the number of communities without risking an arbitrary number, we compared the choices of two probabilistic methods, the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). Since the communities are updated periodically, we also analyzed the impact of aged samples on the proposed framework. Finally, we compared the performance of our FL-based solution to a centralized training approach. We analyzed similarity and dissimilarity metrics on mobility samples and the contact time of individuals in three different scenarios. Our results indicate that AIC outperforms BIC when choosing the number of communities, although both satisfy the evaluation metrics. We also found that using older samples benefits more complex spatio-temporal scenarios. Finally, no significant losses were detected when compared to a centralized training approach, reinforcing the advantages of using the FL-based method.</p></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":null,"pages":null},"PeriodicalIF":4.4000,"publicationDate":"2024-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Ad Hoc Networks","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1570870524002002","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Identifying individuals with similar behaviors and mobility patterns has become essential to improving the functioning of urban services. However, since these patterns can vary over time, such identification needs to be done periodically. Furthermore, once mobility data expresses the routine of individuals, privacy must be guaranteed. In this work, we propose a framework for periodically detecting and grouping individuals with behavioral similarities into communities. To accomplish this, we built an autoencoder model to extract spatio-temporal mobility features from raw user data at periodic intervals. We used Federated Learning (FL) as a training approach to preserve privacy and alleviate time-consuming training and communication costs. To determine the number of communities without risking an arbitrary number, we compared the choices of two probabilistic methods, the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). Since the communities are updated periodically, we also analyzed the impact of aged samples on the proposed framework. Finally, we compared the performance of our FL-based solution to a centralized training approach. We analyzed similarity and dissimilarity metrics on mobility samples and the contact time of individuals in three different scenarios. Our results indicate that AIC outperforms BIC when choosing the number of communities, although both satisfy the evaluation metrics. We also found that using older samples benefits more complex spatio-temporal scenarios. Finally, no significant losses were detected when compared to a centralized training approach, reinforcing the advantages of using the FL-based method.
期刊介绍:
The Ad Hoc Networks is an international and archival journal providing a publication vehicle for complete coverage of all topics of interest to those involved in ad hoc and sensor networking areas. The Ad Hoc Networks considers original, high quality and unpublished contributions addressing all aspects of ad hoc and sensor networks. Specific areas of interest include, but are not limited to:
Mobile and Wireless Ad Hoc Networks
Sensor Networks
Wireless Local and Personal Area Networks
Home Networks
Ad Hoc Networks of Autonomous Intelligent Systems
Novel Architectures for Ad Hoc and Sensor Networks
Self-organizing Network Architectures and Protocols
Transport Layer Protocols
Routing protocols (unicast, multicast, geocast, etc.)
Media Access Control Techniques
Error Control Schemes
Power-Aware, Low-Power and Energy-Efficient Designs
Synchronization and Scheduling Issues
Mobility Management
Mobility-Tolerant Communication Protocols
Location Tracking and Location-based Services
Resource and Information Management
Security and Fault-Tolerance Issues
Hardware and Software Platforms, Systems, and Testbeds
Experimental and Prototype Results
Quality-of-Service Issues
Cross-Layer Interactions
Scalability Issues
Performance Analysis and Simulation of Protocols.