华电技术 ›› 2021, Vol. 43 ›› Issue (2): 15-21.doi: 10.3969/j.issn.1674-1951.2021.02.003

• 电力数据安全 • 上一篇    下一篇

G3-PLC中MAC层动态信任评价机制研究

董重重1(), 谢玮1, 孙秉宇1, 简子倪2, 蒋究2,*(), 王先培2()   

  1. 1.国网湖北省电力公司计量中心,武汉 430080
    2.武汉大学 电子信息学院,武汉 430072
  • 收稿日期:2020-07-10 修回日期:2021-01-25 出版日期:2021-02-25 发布日期:2021-03-05
  • 通讯作者: 蒋究
  • 作者简介:董重重(1985—),男,湖北武汉人,工程师,硕士,从事软件工程、调度自动化工作(E-mail: 414396666@qq.com)|王先培(1962—),男,湖北武汉人,教授,博士生导师,从事电力系统可靠性分析、系统集成与故障诊断等方面的研究(E-mail: xpwang@whu.edu.cn)。
  • 基金资助:
    国家自然科学基金项目(51707135)

Research on dynamic trust evaluation mechanism on MAC layer in G3-PLC

DONG Chongchong1(), XIE Wei1, SUN Bingyu1, JIAN Zini2, JIANG Jiu2,*(), WANG Xianpei2()   

  1. 1. State Grid Hubei Electric Power Company Metering Center,Wuhan 430080,China
    2. Electronic Information School,Wuhan University, Wuhan 430072,China
  • Received:2020-07-10 Revised:2021-01-25 Online:2021-02-25 Published:2021-03-05
  • Contact: JIANG Jiu

摘要:

G3-PLC是目前应用最为广泛的电力线通信标准协议之一,适用于智能电网的抄表、监测等传输数据速度要求不高的场合。为保障G3标准下电力线载波通信安全,以无线通信安全技术标准为依据,在媒体接入控制(MAC)层已有的MAC通信层和6LoWPAN安全认证层的基础上增加MAC感知子层,并建立动态信任评价机制模型,实现对用户频谱资源的合理分配。该信任机制基于客户端用户行为进行评价,利用强化学习方法,采用信任值升降作为奖惩制度,以信任值为判据实现服务端对客户端频谱资源的分配,确保每个感知用户在通信过程中与信誉高的节点融合并降低恶意用户的信任值,从而进行正确频谱感知,抑制分配不信任用户频谱资源,最终将不信任用户剔除出整个网络。

关键词: G3-PLC, 无线通信, 智能电网, 媒体接入控制层, 频谱资源, 信任评价机制, 强化学习, 客户端, 服务端

Abstract:

G3-PLC is one of the most widely used power line communication standard protocols. It is suitable for applications which make moderate requests on data transmission speed, such as meter reading and smart grid monitoring. In order to ensure the security of power line carrier communication under the standard G3, an extra MAC sensing sublayer was added on the existing MAC communication layer and 6LoWPAN security authentication layer of Media Access Control (MAC) layer, in compliance with the wireless communication security technology standard. And a dynamic trust evaluation mechanism model was constructed to make reasonable allocation of user spectrum resources. The trust mechanism makes evaluation according to users' behavior by reinforcement learning. Taking trust value as the reward and punishment system, the trust value is used as the criterion to distribute spectrum resources between clients. Ensuring that each perception client can be fused with high-reputation nodes in the process of communication,the trust value of malicious users will be reduced. The method can perform correct spectrum sensing, suppress the spectrum resources allocated to malicious users, and finally remove malicious users away from the entire network.

Key words: G3-PLC, wireless communication, smart grid, MAC layer, spectrum resources, trust evaluation model, reinforcement learning, client, server

中图分类号: