Scholarlink

<p>Network function virtualization (NFV) and mobile edge computing (MEC) have been introduced into the Internet of Things (IoT) to meet diverse user requests. However, the frequent data transmission between edge servers makes the network vulnerable to distributed denial of service (DDoS) attacks, allowing attackers to occupy the server’s available resources and threaten user service security deployment. To provide users with secure and reliable services, we have constructed the optimization model and introduced the detection of malicious devices based on Consortium Blockchain (CB). Then, we have proposed an attack-aware SFC deployment algorithm based on deep reinforcement learning (DRL) and CB, abbreviated as the CB-DRL, which leverages CB to detect and mitigate attacks and a sequence-to-sequence model consisting of multiple long short-term memory networks (LSTMs) to deploy SFC. Compared to other algorithms, the experimental results show that the CB-DRL can improve the load-balancing level and request acceptance ratio by 6.16% and 6.61%, 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

End-to-End Attack-Aware Service Function Chains Deployment Based on Consortium Blockchain in IoT

  • 期刊论文

  • 发表日期：

    2025年9月6日

查看全文
  • Yanrui Song,
  • Lei Zhuang,
  • Zexi Xu,
  • Guoqing Wang,
  • Xu Feng,
  • Wenshuai Mo

摘要

Network function virtualization (NFV) and mobile edge computing (MEC) have been introduced into the Internet of Things (IoT) to meet diverse user requests. However, the frequent data transmission between edge servers makes the network vulnerable to distributed denial of service (DDoS) attacks, allowing attackers to occupy the server’s available resources and threaten user service security deployment. To provide users with secure and reliable services, we have constructed the optimization model and introduced the detection of malicious devices based on Consortium Blockchain (CB). Then, we have proposed an attack-aware SFC deployment algorithm based on deep reinforcement learning (DRL) and CB, abbreviated as the CB-DRL, which leverages CB to detect and mitigate attacks and a sequence-to-sequence model consisting of multiple long short-term memory networks (LSTMs) to deploy SFC. Compared to other algorithms, the experimental results show that the CB-DRL can improve the load-balancing level and request acceptance ratio by 6.16% and 6.61%, respectively.