Classification and Analysis of Adversarial Machine Learning Attacks in IoT: a Label Flipping Attack Case Study

Abrishami, Mahdi; Dadkhah, Sajjad; Neto, Euclides Carlos Pinto; Xiong, Pulei; Iqbal, Shahrear; Ray, Suprio; Ghorbani, Ali A.

doi:10.23919/fruct56874.2022.9953823

Cited by 3 publications

(1 citation statement)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After the big growth in IoT applications and devices and after being surrounded by varies vulnerabilities, it becomes harder to secure those systems with humans, ML techniques play an important role in regular and IoT networks security. Considering the various vulnerabilities in the IoT domain, ML algorithms are widely being used to tackle the potential Issues [226]. The requirements for securing IoT devices have become complex because several technologies, from physical devices and wireless transmission to mobile and cloud architectures, need to be secured and combined with other technologies.…”

Section: And DL In Iot Securitymentioning

confidence: 99%

A Survey of Using Machine Learning in IoT Security and the Challenges Faced by Researchers

Harahsheh,

Chen

2023

IJCAI

View full text Add to dashboard Cite

The Internet of Things (IoT) has become more popular in the last 15 years as it has significantly improved and gained control in multiple fields. We are nowadays surrounded by billions of IoT devices that directly integrate with our lives, some of them are at the center of our homes, and others control sensitive data such as military fields, healthcare, and datacenters, among others. This popularity makes factories and companies compete to produce and develop many types of those devices without caring about how secure they are. On the other hand, IoT is considered a good insecure environment for cyber thefts. Machine Learning (ML) and Deep Learning (DL) also gained more importance in the last 15 years; they achieved success in the networking security field too. IoT has some similar security requirements such as traditional networks, but with some differences according to its characteristics, some specific security features, and environmental limitations, some differences are made such as low energy resources, limited computational capability, and small memory. These limitations inspire some researchers to search for the perfect and lightweight security ways which strike a balance between performance and security. This survey provides a comprehensive discussion about using machine learning and deep learning in IoT devices within the last five years. It also lists the challenges faced by each model and algorithm. In addition, this survey shows some of the current solutions and other future directions and suggestions. It also focuses on the research that took the IoT environment limitations into consideration.Povzetek: Podan je pregled uporabe strojnega in globokega učenja v IoT napravah ter izzivov, rešitev in prihodnjih smeri raziskav.Informatica 47 (2023) 1-54 k. Harahsheh et al.

show abstract

Section: And DL In Iot Securitymentioning

confidence: 99%

A Survey of Using Machine Learning in IoT Security and the Challenges Faced by Researchers

Harahsheh,

Chen

2023

IJCAI

View full text Add to dashboard Cite

show abstract

Privacy-Preservation Robust Federated Learning with Blockchain-based Hierarchical Framework

Liu,

Li,

Chen

et al. 2024

International Conference on Computing, Machine Learning and Data Science

View full text Add to dashboard Cite

Mitigating Adversarial Attacks against IoT Profiling

Neto,

Dadkhah,

Sadeghi

et al. 2024

Electronics

View full text Add to dashboard Cite

Internet of Things (IoT) applications have been helping society in several ways. However, challenges still must be faced to enable efficient and secure IoT operations. In this context, IoT profiling refers to the service of identifying and classifying IoT devices’ behavior based on different features using different approaches (e.g., Deep Learning). Data poisoning and adversarial attacks are challenging to detect and mitigate and can degrade the performance of a trained model. Thereupon, the main goal of this research is to propose the Overlapping Label Recovery (OLR) framework to mitigate the effects of label-flipping attacks in Deep-Learning-based IoT profiling. OLR uses Random Forests (RF) as underlying cleaners to recover labels. After that, the dataset is re-evaluated and new labels are produced to minimize the impact of label flipping. OLR can be configured using different hyperparameters and we investigate how different values can improve the recovery procedure. The results obtained by evaluating Deep Learning (DL) models using a poisoned version of the CIC IoT Dataset 2022 demonstrate that training overlap needs to be controlled to maintain good performance and that the proposed strategy improves the overall profiling performance in all cases investigated.

show abstract

Classification and Analysis of Adversarial Machine Learning Attacks in IoT: a Label Flipping Attack Case Study

Cited by 3 publications

References 50 publications

A Survey of Using Machine Learning in IoT Security and the Challenges Faced by Researchers

A Survey of Using Machine Learning in IoT Security and the Challenges Faced by Researchers

Privacy-Preservation Robust Federated Learning with Blockchain-based Hierarchical Framework

Mitigating Adversarial Attacks against IoT Profiling

Contact Info

Product

Resources

About