Toward IoT Node Authentication Mechanism in Next Generation Networks

Nguyen, Dinh Duc Nha; Sood, Keshav; Xiang, Yong; Gao, Longxiang; Chi, Lianhua; Yu, Shui

doi:10.1109/jiot.2023.3262822

Cited by 9 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [169], the authors present a Mahalanobis distance and Chi-squared distribution RF fingerprinting approach focused on providing SEI-based authentication within 5G IoT next-generation networks. The authors show that their approach requires lower training times and fewer resources than five other SEI processes while achieving a higher average accuracy.…”

Section: Sei On Resource-constrained Devicesmentioning

confidence: 99%

Considerations, Advances, and Challenges Associated with the Use of Specific Emitter Identification in the Security of Internet of Things Deployments: A Survey

Tyler,

Fadul,

Reising

2023

Information

View full text Add to dashboard Cite

Initially introduced almost thirty years ago for the express purpose of providing electronic warfare systems the capabilities to detect, characterize, and identify radar emitters, Specific Emitter Identification (SEI) has recently received a lot of attention within the research community as a physical layer technique for securing Internet of Things (IoT) deployments. This attention is largely due to SEI’s demonstrated success in passively and uniquely identifying wireless emitters using traditional machine learning and the success of Deep Learning (DL) within the natural language processing and computer vision areas. SEI exploits distinct and unintentional features present within an emitter’s transmitted signals. These distinctive and unintentional features are attributed to slight manufacturing and assembly variations within and between the components, sub-systems, and systems comprising an emitter’s Radio Frequency (RF) front end. Although sufficient to facilitate SEI, these features do not hinder normal operations such as detection, channel estimation, timing, and demodulation. However, despite the plethora of SEI publications, it has remained largely a focus of academic endeavors, primarily focusing on proof-of-concept demonstration and little to no use in operational networks for various reasons. The focus of this survey is a review of SEI publications from the perspective of its use as a practical, effective, and usable IoT security mechanism; thus, we use IoT requirements and constraints (e.g., wireless standard, nature of their deployment) as a lens through which each reviewed paper is analyzed. Previous surveys have not taken such an approach and have only used IoT as motivation, a setting, or a context. In this survey, we consider operating conditions, SEI threats, SEI at scale, publicly available data sets, and SEI considerations that are dictated by the fact that it is to be employed by IoT devices or IoT infrastructure.

show abstract

Section: Sei On Resource-constrained Devicesmentioning

confidence: 99%

Considerations, Advances, and Challenges Associated with the Use of Specific Emitter Identification in the Security of Internet of Things Deployments: A Survey

Tyler,

Fadul,

Reising

2023

Information

View full text Add to dashboard Cite

show abstract

“…Extraction of features is sometimes timeconsuming and difficult because previous machine learning algorithms mainly rely on feature engineering. Therefore, using typical machine learning techniques like DT [27], ensemble models [28], ANN [29], AE [30], RNN [31], KNN [3], and DNN [32] in real-time applications is impractical for detecting threats. Table 1 presents the review of various IDS methodologies used for protecting 5G-IoT networks.…”

Section: Related Workmentioning

confidence: 99%

“…Internet of Things (IoT) is more susceptible to the different types of intrusions or attacks, due to its tremendous growth in present days [1,2]. The effectiveness of sensitive devices may affect end users, increases cyber threats and identity theft, which also has an adverse effect on income since problems produced by IoT network are unnoticed for extended periods of time [3][4][5]. Moreover, attacks on IoT interfaces must be closely monitored in real time for ensuring safety and security.…”

Section: Introductionmentioning

confidence: 99%

An Avant-Garde African Vulture Optimization (A2VO) based Deep RNN-LSTM Model for 5G-IoT Security

Gopalakrishnan Ramasubramanian,

Singaravelu Rajaprakash

2023

ARASET

View full text Add to dashboard Cite

In current days, 5G is more essential for the Internet of Things (IoT) systems, since it offers a quicker network with more capacity to address communication needs. The frequency range that mobile communication technologies employ to transfer data is expanded by the 5G spectrum. However, it also broadens the presence of attacks within the core network, increasing the corporate system's susceptibility to security intrusions. Hence, various Intrusion Detection System (IDS) frameworks are developed in the conventional works for 5G-IoT network security. However, the significant challenges of the existing studies are complexity burden, high processing time, and ineffective intrusion detection. Therefore, the proposed work intends to implement a novel Avant-Garde African Vulture Optimization (A2VO), and Recurrent Neural Network (RNN) – Long Short-Term Memory (LSTM) mechanism of 5G-IoT security. The min-max normalization technique is used to preprocess the attributes and fields in the dataset. In addition, a precise and effective deep Recurrent Neural Network (RNN) - Long Short Term Memory (LSTM) based classification algorithm is used to categorize normal and attacking data with high accuracy and low complication. The most significant features from the normalized dataset are chosen for classification using the Avant-Garde African Vulture Optimization (A2VO) method. For demonstrating the superiority of the suggested security model, the performance results are verified and compared using several benchmarking datasets.

show abstract