Scholarlink

<p>The developments of Internet of Things (IoT), smart cities have become majority of urbanization. IoT networks use the Internet as an open channel to enable distributed smart devices to collect and process data within the architecture of smart cities. In this paper, a privacy-preserving and secure framework using a blockchain-based quantum-inspired complex convolutional neural network for IoT-driven smart cities (PSF-BCH-QICCN-IoT) is proposed. Initially, the dataset was taken from the BoT-IoT dataset. The data collected are fed to blockchain-based Proof-of-Monitoring (PoM) for privacy-preserving and secure framework. Then feature mapping and feature selection are formed using the Hunger Game Search Optimization Algorithm (HGSOA). After that, QICCN is utilized to classify anomalies such as Denial-of-Service, Distributed DoS, Normal, Reconnaissance, and Theft. Generally, QICCN does not show some optimization adaption techniques to determine the optimum parameter to offer accurate detection. Firebug Swarm Optimization (FSO) process is proposed to enhance QICCN and classify the anomaly precisely. The performance of proposed technique is analyzed using performance metrics such as accuracy, specificity, recall, precision, F1-score, false alarm rate. The proposed (PSF-BCH-QICCN-IoT) method achieves accuracies of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\mathbf {88.67\%, 87.23\%}\)</EquationSource> </InlineEquation>, and <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\mathbf {90.45\%}\)</EquationSource> </InlineEquation> on the respective datasets, which correspond to improvements of <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(\mathbf {23.33\%, 21.45\%}\)</EquationSource> </InlineEquation>, and <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(\mathbf {31.35\%}\)</EquationSource> </InlineEquation> over the baseline accuracies of <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(\mathbf {71.87\%, 71.82\%}\)</EquationSource> </InlineEquation>, and <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(\mathbf {68.84\%}\)</EquationSource> </InlineEquation>; <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(\mathbf {34.15\%, 32.26\%}\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\(\mathbf {19.95\%}\)</EquationSource> </InlineEquation> higher precision; <InlineEquation ID="IEq9"> <EquationSource Format="TEX">\(\mathbf {25.55\%, 27.35\%}\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq10"> <EquationSource Format="TEX">\(\mathbf {22.15\%}\)</EquationSource> </InlineEquation> higher recall analyzed to the existing methods, like developing effectual feature engineering along machine learning technique for identifying IoT-botnet cyber-attacks (DMLP-IoT-BAD), feature engineering depend performance analysis of ML-DL processes for Botnet attack identification in IoMT (SVM–IoT-BAD) and intrusion identification scheme for IoT botnet attacks utilizing deep learning (DNN-IoT-BAD), respectively.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A privacy-preserving and secure framework using blockchain-based quantum-inspired complex convolutional neural network for IoT-driven smart cities

  • 期刊论文

  • 发表日期：

    2025年11月11日

查看全文
  • Chandra Prakash Singh,
  • Rohita Yamaganti,
  • Lokendra Singh Umrao

摘要

The developments of Internet of Things (IoT), smart cities have become majority of urbanization. IoT networks use the Internet as an open channel to enable distributed smart devices to collect and process data within the architecture of smart cities. In this paper, a privacy-preserving and secure framework using a blockchain-based quantum-inspired complex convolutional neural network for IoT-driven smart cities (PSF-BCH-QICCN-IoT) is proposed. Initially, the dataset was taken from the BoT-IoT dataset. The data collected are fed to blockchain-based Proof-of-Monitoring (PoM) for privacy-preserving and secure framework. Then feature mapping and feature selection are formed using the Hunger Game Search Optimization Algorithm (HGSOA). After that, QICCN is utilized to classify anomalies such as Denial-of-Service, Distributed DoS, Normal, Reconnaissance, and Theft. Generally, QICCN does not show some optimization adaption techniques to determine the optimum parameter to offer accurate detection. Firebug Swarm Optimization (FSO) process is proposed to enhance QICCN and classify the anomaly precisely. The performance of proposed technique is analyzed using performance metrics such as accuracy, specificity, recall, precision, F1-score, false alarm rate. The proposed (PSF-BCH-QICCN-IoT) method achieves accuracies of \(\mathbf {88.67\%, 87.23\%}\) , and \(\mathbf {90.45\%}\) on the respective datasets, which correspond to improvements of \(\mathbf {23.33\%, 21.45\%}\) , and \(\mathbf {31.35\%}\) over the baseline accuracies of \(\mathbf {71.87\%, 71.82\%}\) , and \(\mathbf {68.84\%}\) ; \(\mathbf {34.15\%, 32.26\%}\) and \(\mathbf {19.95\%}\) higher precision; \(\mathbf {25.55\%, 27.35\%}\) and \(\mathbf {22.15\%}\) higher recall analyzed to the existing methods, like developing effectual feature engineering along machine learning technique for identifying IoT-botnet cyber-attacks (DMLP-IoT-BAD), feature engineering depend performance analysis of ML-DL processes for Botnet attack identification in IoMT (SVM–IoT-BAD) and intrusion identification scheme for IoT botnet attacks utilizing deep learning (DNN-IoT-BAD), respectively.