Jove
Visualize
Contact Us

Related Experiment Video

Updated: Nov 29, 2025

Data Communication Based on MQTT in a Polymer Extrusion Process
08:15

Data Communication Based on MQTT in a Polymer Extrusion Process

Published on: July 15, 2022

3.7K

MQTTset, a New Dataset for Machine Learning Techniques on MQTT.

Ivan Vaccari1,2, Giovanni Chiola2, Maurizio Aiello1

  • 1Consiglio Nazionale delle Ricerche (CNR), IEIIT Institute, 16149 Genoa, Italy.

Sensors (Basel, Switzerland)
|November 21, 2020
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Dual-View Single-Shot Multibox Detector at Urban Intersections: Settings and Performance Evaluation.

Sensors (Basel, Switzerland)·2023
Same author

Preventing MQTT Vulnerabilities Using IoT-Enabled Intrusion Detection System.

Sensors (Basel, Switzerland)·2022
Same author

Identifying and Mitigating Phishing Attack Threats in IoT Use Cases Using a Threat Modelling Approach.

Sensors (Basel, Switzerland)·2021
Same author

A Generative Adversarial Network (GAN) Technique for Internet of Medical Things Data.

Sensors (Basel, Switzerland)·2021
Same author

SlowITe, a Novel Denial of Service Attack Affecting MQTT.

Sensors (Basel, Switzerland)·2020
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

This article introduces MQTTset, a specialized dataset designed to help researchers train machine learning models for detecting cyber-attacks within IoT networks that use the MQTT communication protocol. The authors demonstrate the dataset's utility by validating it against simulated malicious activity, showing that it effectively supports the development of security systems for connected environments.

Area of Science:

  • Cybersecurity research within MQTTset network protection
  • Machine learning applications in IoT security

Background:

Modern internet of things networks frequently monitor sensitive environments, leading to a massive surge in transmitted information. That uncertainty drove researchers to prioritize the protection of these interconnected systems against potential threats. Prior research has shown that safeguarding numerous devices remains a significant challenge for modern digital infrastructure. Detection systems utilizing advanced computational algorithms represent a primary defense strategy in current cybersecurity efforts. No prior work had resolved the scarcity of specialized training data tailored specifically for these communication protocols. This gap motivated the development of resources that accurately reflect the unique traffic patterns found in such environments. Existing security frameworks often lack the specific information necessary to identify sophisticated malicious activities effectively. Consequently, the field requires robust benchmarks to improve the performance of automated threat identification tools.

Purpose Of The Study:

Keywords:
Internet of ThingsMQTTartificial intelligencedatasetdetection systemmachine learningIoT securitynetwork traffic analysisthreat detectionmachine learning models

Frequently Asked Questions

The researchers propose that the dataset enables machine learning models to identify cyber-attacks by training on a combination of legitimate traffic and simulated malicious activity, which improves the detection accuracy for threats targeting the specific communication protocol.

The authors utilize the Message Queuing Telemetry Transport protocol, which serves as the foundational communication standard for the dataset, allowing for the simulation of realistic network traffic patterns within the internet of things environment.

A validation system is necessary to confirm the utility of the data, as it allows the authors to demonstrate that the combined legitimate and malicious traffic can effectively train models to distinguish between normal and harmful network behaviors.

Related Experiment Videos

Last Updated: Nov 29, 2025

Data Communication Based on MQTT in a Polymer Extrusion Process
08:15

Data Communication Based on MQTT in a Polymer Extrusion Process

Published on: July 15, 2022

3.7K

The aim of this study is to introduce a specialized dataset designed to support machine learning techniques for securing communication protocols. Researchers addressed the lack of specific training data for identifying threats in connected environments. They sought to create a resource that accurately represents the traffic patterns of the protocol in question. The authors intended to validate the utility of this information through a practical detection system. This motivation stemmed from the increasing prevalence of cyber-attacks targeting interconnected devices. By providing this resource, the team hopes to facilitate the development of more effective security models. They focused on the specific requirements of modern networks to ensure the data remains relevant. The study addresses the urgent need for tools that protect critical infrastructure from malicious activity.

Main Methods:

The review approach involved constructing a comprehensive collection of network traffic data focused on a specific communication standard. Researchers generated legitimate traffic logs to establish a baseline for normal operational behavior. They integrated various cyber-attacks into this baseline to simulate realistic malicious scenarios within the network. The team employed machine learning techniques to process and analyze the resulting information. Validation occurred through the implementation of a hypothetical detection system designed to identify these injected threats. This approach ensured the data could support the training of predictive security models. The methodology prioritized the inclusion of diverse attack vectors to enhance the robustness of the final resource. Investigators carefully curated the information to reflect the complexities of modern connected environments.

Main Results:

Key findings from the literature indicate that the generated resource successfully supports the training of machine learning models for threat identification. The authors demonstrate that their validation system effectively distinguishes between normal traffic and malicious activity. This result confirms the practical utility of the dataset for enhancing security in connected environments. The study shows that integrating specific attack scenarios provides a necessary benchmark for evaluating detection performance. Researchers observed that models trained on this information achieved reliable identification of threats targeting the protocol. The findings highlight the importance of protocol-specific data in developing effective defense mechanisms. The evidence suggests that the dataset serves as a functional tool for researchers building automated security solutions. These results validate the approach of combining legitimate and malicious data for training purposes.

Conclusions:

The authors suggest that their new resource provides a reliable foundation for training automated security models. This work demonstrates that combining legitimate traffic with simulated threats creates a viable testing environment. The researchers propose that their approach effectively supports the development of protective systems for connected infrastructures. Their findings indicate that machine learning models trained on this data can successfully identify various cyber-attacks. The study highlights the utility of protocol-specific information in enhancing network defense capabilities. The authors claim that their validation process confirms the practical applicability of the generated information. This synthesis implies that specialized datasets are necessary for advancing security in specific communication architectures. The researchers conclude that their contribution facilitates the creation of more resilient digital environments.

The dataset functions as a training resource, providing the labeled information required for machine learning algorithms to learn the distinct characteristics of both standard operations and various cyber-attacks within the network.

The researchers measure the effectiveness of their resource by evaluating the performance of a detection system trained on the data, specifically observing its capacity to identify simulated malicious events against the network.

The authors propose that their work provides a pathway for implementing robust security measures in connected contexts, suggesting that specialized data is a prerequisite for protecting modern digital infrastructures from evolving threats.