Affiliation:
1. Escuela de Ingeniería en Ciberseguridad, Facultad de Ingenierías Ciencias Aplicadas, Universidad de Las Américas, Quito 170125, Ecuador
Abstract
This study addresses anomaly detection in smart city environments driven by the Internet of Things. In these cities, digital interconnection and the extensive network of sensors generate enormous amounts of data, which are essential to improving citizens’ efficiency and quality of life. However, this data may also contain strange events that require early detection to ensure the proper functioning of urban systems. For this, anomaly detection models are explored to identify unusual patterns in urban data. The work focuses on the applicability and effectiveness of these models in different urban scenarios supported by the Internet of Things. Furthermore, its performance is evaluated by comparing it with existing approaches, and its advantages and limitations are analyzed. The results show that the proposed models, including Isolation Forest, recurrent neural network, and variational autoencoder, are highly effective in detecting anomalies in urban data. This work contributes to the field of smart cities by improving the safety and efficiency of urban systems. Early detection of anomalies makes it possible to prevent unplanned interruptions, ensure the safety of citizens, and maintain the integrity of urban systems. Furthermore, the relevance of this work in the existing literature and its importance for the evolution of smart cities supported by the Internet of Things are highlighted.
Subject
Fluid Flow and Transfer Processes,Computer Science Applications,Process Chemistry and Technology,General Engineering,Instrumentation,General Materials Science
Reference48 articles.
1. Dense Air Quality Sensor Networks: Validation, Analysis, and Benefits;Zaidan;IEEE Sens. J.,2022
2. Sequential Attention Mechanism for Weakly Supervised Video Anomaly Detection;Ullah;Expert. Syst. Appl.,2023
3. Shukla, S., Thakur, S., and Breslin, J.G. (2022). Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer.
4. A Trust Aware Security Mechanism to Detect Sinkhole Attack in RPL-Based IoT Environment Using Random Forest—RFTRUST;Prathapchandran;Comput. Netw.,2021
5. Unfolding the Network Dataset to Understand the Contribution of Features for Detecting Malicious Activities Using AI/ML;Khatkar;Mater. Today Proc.,2022