Scholarlink

<p>The integration of wearable and implantable medical devices (WIMDs) within Internet of Things (IoT)-based healthcare systems enables real-time health monitoring but introduces significant data security challenges. While blockchain technology ensures transparency and traceability, the lack of lightweight authentication mechanisms for resource-constrained WIMDs remains a critical issue. To address this, a secure authentication framework using blockchain and Bayesian regularization back propagation neural network for the protection of wearable and implantable medical devices in IoT-based healthcare environments (SA-BRBPNN-WIMI-IoT) is proposed. The objective of this paper is to develop a secure, efficient, and scalable authentication system ensuring data confidentiality, integrity, and authorized access. Initially, ECG signals from the physikalisch-technische bundesanstalt—extra large electrocardiography dataset are pre-processed using learnable edge collaborative filtering to eliminate noise and artifacts. Scalable affine multi-view subspace clustering segments the denoised signals to isolate individual heartbeats. Deep wavelet scattering transform extracts discriminative features, which are binarized and used for cryptographic key generation via Bayesian regularization back propagation neural network (BRBPNN), producing high-entropy, user-specific keys. To secure medical data transmission, low-complexity elliptic galois cryptography provides lightweight encryption and decryption. Blockchain with a proof-of-reputation consensus mechanism employed for decentralized access control and tamper-proof data logging. The proposed SA-BRBPNN-WIMI-IoT is implemented and the performance metrics, such as accuracy, false acceptance rate (FAR), false rejection rate (FRR), entropy, success rate, encryption time and decryption time are analyzed. The performance of proposed method attains 99.56% higher accuracy, 0.25% lower FAR and 0.31% lower FRR while compared with existing methods 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

An Advanced Security Authentication Integrating Blockchain and Bayesian Regularization Back Propagation Neural Network for Protection of Wearable and Implantable Medical Devices in IoT Based Healthcare Environment

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • G. Arun Jeba Kumar,
  • V. Evelyn Brindha

摘要

The integration of wearable and implantable medical devices (WIMDs) within Internet of Things (IoT)-based healthcare systems enables real-time health monitoring but introduces significant data security challenges. While blockchain technology ensures transparency and traceability, the lack of lightweight authentication mechanisms for resource-constrained WIMDs remains a critical issue. To address this, a secure authentication framework using blockchain and Bayesian regularization back propagation neural network for the protection of wearable and implantable medical devices in IoT-based healthcare environments (SA-BRBPNN-WIMI-IoT) is proposed. The objective of this paper is to develop a secure, efficient, and scalable authentication system ensuring data confidentiality, integrity, and authorized access. Initially, ECG signals from the physikalisch-technische bundesanstalt—extra large electrocardiography dataset are pre-processed using learnable edge collaborative filtering to eliminate noise and artifacts. Scalable affine multi-view subspace clustering segments the denoised signals to isolate individual heartbeats. Deep wavelet scattering transform extracts discriminative features, which are binarized and used for cryptographic key generation via Bayesian regularization back propagation neural network (BRBPNN), producing high-entropy, user-specific keys. To secure medical data transmission, low-complexity elliptic galois cryptography provides lightweight encryption and decryption. Blockchain with a proof-of-reputation consensus mechanism employed for decentralized access control and tamper-proof data logging. The proposed SA-BRBPNN-WIMI-IoT is implemented and the performance metrics, such as accuracy, false acceptance rate (FAR), false rejection rate (FRR), entropy, success rate, encryption time and decryption time are analyzed. The performance of proposed method attains 99.56% higher accuracy, 0.25% lower FAR and 0.31% lower FRR while compared with existing methods respectively.