电子学报 ›› 2016, Vol. 44 ›› Issue (7): 1555-1560.DOI: 10.3969/j.issn.0372-2112.2016.07.005

• 学术论文 • 上一篇    下一篇

抗脉冲干扰的分布式仿射投影符号算法

倪锦根, 马兰申   

  1. 苏州大学电子信息学院, 江苏苏州 215006
  • 收稿日期:2015-06-29 修回日期:2015-09-30 出版日期:2016-07-25
    • 作者简介:
    • 倪锦根 男,1979年11月生,江苏省兴化市人.毕业于复旦大学,获理学博士学位,现为苏州大学电子信息学院副教授、硕士生导师,IEEE会员.主要研究方向为自适应滤波、自适应网络、多采样率滤波器组设计、高速数字系统的FPGA实现.E-mail:jni@suda.edu.cn;马兰申 男,1988年4月生,安徽省亳州市人.获苏州大学硕士学位.研究方向为自适应滤波和分布式估计.E-mail:malanlong@163.com
    • 基金资助:
    • 国家自然科学基金 (No.61471251,No.61101217); 江苏省自然科学基金 (No.BK20131164)

Distributed Affine Projection Sign Algorithms Against Impulsive Interferences

NI Jin-gen, MA Lan-shen   

  1. School of Electronic and Information Engineering, Soochow University, Suzhou, Jiangsu 215006, China
  • Received:2015-06-29 Revised:2015-09-30 Online:2016-07-25 Published:2016-07-25
    • Supported by:
    • National Natural Science Foundation of China (No.61471251, No.61101217); Natural Science Foundation of Jiangsu Province,  China (No.BK20131164)

摘要:

递增式和扩散式仿射投影算法收敛较快,但在脉冲噪声环境下这两种分布式估计算法收敛性较差或容易发散.本文采用受网络节点的权值向量更新约束的后验误差向量?1范数最小化方法,提出了两种抗脉冲干扰的分布式估计算法,即递增式和扩散式仿射投影符号算法.仿真结果表明,与分布式仿射投影算法相比,分布式仿射投影符号算法在脉冲噪声环境下具有更好的鲁棒性.

关键词: 自适应网络, 仿射投影, 符号算法, 分布式估计

Abstract:

While the incremental and diffusion affine projection algorithms have fast convergence rate,they may suffer from bad convergence performance or divergence in impulsive interference environments. Based on the method of minimization of the ?1-norm of the a posteriori error vectors subject to a constraint on the update of the weight vectors of nodes in the network,two distributed estimation algorithms against impulsive interferences are proposed,namely,the incremental and diffusion affine projection sign algorithms. Simulation results show that the distributed affine projection sign algorithms exhibit better robustness than the distributed affine projection algorithms in impulsive interference environments.

Key words: adaptive network, affine projection, sign algorithm, distributed estimation

中图分类号: