An optimized hybrid deep neural network architecture for intrusion detection in real-time IoT networks,Transactions on Emerging Telecommunications Technologies

当前位置： X-MOL 学术 › Trans. Emerg. Telecommun. Technol. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

An optimized hybrid deep neural network architecture for intrusion detection in real-time IoT networks
Transactions on Emerging Telecommunications Technologies ( IF 2.5 ) Pub Date : 2022-07-29 , DOI: 10.1002/ett.4609
M. Shobana ₁ , C. Shanmuganathan ₂ , Nagendra Panini Challa ₃ , S. Ramya ₄

Affiliation

The Internet-of-Things (IoT) refers to the interconnection of things to the physical network that is embedded with software, sensors, and other devices to exchange information from one device to the other. The interconnection of devices means there is the possibility of challenges like security, trustworthiness, reliability, confidentiality, and so on. To address those issues, we have proposed a novel GTBSS-HDNN approach which hybridization of Group theory (GT), Binary Spring search (BSS) algorithm, and Hybrid deep neural network (HDNN). The proposed GTBSS-HDNN approach effectively detects the intrusion in the IoT nodes. Initially, the privacy-preserving technology was implemented using a Blockchain-based methodology. Our proposed privacy-preserving methods are divided into two parts. The first stage utilizes blockchain and the second stage involves Modified Independent Component Algorithm (MICA) to prevent intrusion attacks. The authentication of data is performed by blockchain-based Enhanced Proof of Work (EPoW) and achieves better authentication. Furthermore, the experimental study is carried out using the ToN-IoT dataset, which is used to evaluate the performance of our proposed work. To analyze the performance we have taken the performance metrics such as F1-measure, Detection Rate, Precision, and Accuracy. The performance analyzes depict that the proposed method effectively preserves the accuracy and thereby avert the intrusion attacks. The proposed model achieved 95.3% accuracy, 96.54% precision, 95.23% recall, and 95.67% F-score values on the ToN-IoT dataset and 96.23% accuracy, 95.94% precision, 97.03% recall, and 96.70% F-score results on the BoT-IoT dataset.

中文翻译：

用于实时物联网网络入侵检测的优化混合深度神经网络架构

物联网 (IoT) 是指物与嵌入软件、传感器和其他设备的物理网络互连，以将信息从一个设备交换到另一个设备。设备互联意味着存在安全性、可信性、可靠性、保密性等挑战的可能性。为了解决这些问题，我们提出了一种新的 GTBSS-HDNN 方法，它混合了群论 (GT)、二进制弹簧搜索 (BSS) 算法和混合深度神经网络 (HDNN)。所提出的 GTBSS-HDNN 方法有效地检测了物联网节点中的入侵。最初，隐私保护技术是使用基于区块链的方法实现的。我们提出的隐私保护方法分为两部分。第一阶段利用区块链，第二阶段涉及改进的独立组件算法（MICA）以防止入侵攻击。数据的认证通过基于区块链的增强工作量证明（EPoW）进行，实现了更好的认证。此外，实验研究是使用 ToN-IoT 数据集进行的，该数据集用于评估我们提出的工作的性能。为了分析性能，我们采用了性能指标，例如F 1-measure、检测率、精密度和准确度。性能分析表明，所提出的方法有效地保持了准确性，从而避免了入侵攻击。所提出的模型在 ToN-IoT 数据集上实现了 95.3% 的准确率、96.54% 的准确率、95.23% 的召回率和 95.67%的F值，在BoT-IoT 数据集。

更新日期：2022-07-29

点击分享查看原文

点击收藏

阅读更多本刊最新论文本刊介绍/投稿指南11