系统工程与电子技术 ›› 2025, Vol. 47 ›› Issue (7): 2389-2398.doi: 10.12305/j.issn.1001-506X.2025.07.31

• 通信与网络 • 上一篇    

基于改进CRITIC-GAHP的通信干扰效果在线评估方法

侯文君, 彭闯, 金虎, 雷迎科   

  1. 国防科技大学电子对抗学院, 安徽 合肥 230009
  • 收稿日期:2024-06-12 出版日期:2025-07-16 发布日期:2025-07-22
  • 通讯作者: 金虎
  • 作者简介:侯文君 (2000—), 女, 硕士研究生, 主要研究方向为通信干扰、人工智能
    彭闯 (1994—), 男, 讲师, 博士, 主要研究方向为人工智能、认知无线电
    金虎 (1974—), 男, 教授, 博士, 主要研究方向为智能干扰、信号处理
    雷迎科 (1975—), 男, 教授, 博士, 主要研究方向为人工智能、通信信号处理
  • 基金资助:
    国家自然科学基金(62071479)

Online evaluation method of communication interference effect based on improved CRITIC-GAHP

Wenjun HOU, Chuang PENG, Hu JIN, Yingke LEI   

  1. College of Electronic Countermeasures, National University of Defense Technology, Hefei 230009, China
  • Received:2024-06-12 Online:2025-07-16 Published:2025-07-22
  • Contact: Hu JIN

摘要:

针对通信干扰效果在线评估的无标签样本分类问题, 提出一种基于主客观权重的通信干扰效果在线评估方法。从发送方抗干扰的角度选取多种通信行为参数作为干扰效果评估指标, 分别使用改进的准则重要性相关法(criteria importance through inter-criteria correlation, CRITIC)和群层次分析法(group analytic hierarchy process, GAHP)确定客观和主观权重。通过最小信息熵原理确定组合权重后, 根据各指标参数的变化程度判断干扰效果, 划分干扰效果等级, 实现无标签样本分类。同时, 提出动态更新各指标归一化参数的方法, 使其应对变化的环境。实验结果表明, 提出的方法在指标参数范围变化和小样本情况下仍具有较稳定的性能。

关键词: 通信干扰效果, 在线评估, 无标签样本分类, 主客观权重

Abstract:

For the problem of online evaluation of communication interference effect without labelled samples, a communication interference effect online evaluation method based on subjective and objective weighting is proposed. This method selects various communication behavior parameters from the perspective of anti-interference of the transmitter as interference effect evaluation indicators, and determines the objective and subjective weights of the indicators respectively using the improved criteria importance through inter-criteria correlation (CRITIC) method and the group analytic hierarchy process (GAHP) method. After determining the combination weight by the principle of minimum information entropy, the interference effect is judged by the degree of change of each indicator parameter and divided into interference effect grades, achieving the classification of unlabelled samples. At the same time, a method of dynamically updating the normalization parameters of each indicator is proposed to adapt to changing environments. The experimental results show that the method proposed in this paper has stable performance under the condition of changing indicator parameter ranges and small sample size.

Key words: communication jamming effect, online evaluation, unlabeled sample classification, subjective and objective weighting

中图分类号: