ISSN: 1673-825X    Imprint: Chongqing University of Posts and Telecommunications Journal
Volume listings
  • Current volume
  •  
  • Search by keywords:
  • Search by 
  • Keywords
  • To
  • Journal archive
Related Links
一种改进重采样的粒子滤波盲分离算法
An improved resampling algorithm for blind separation of particle filtering
DOI:10.3979/j.issn.1673-825X.2019.04.010
Received:July 04, 2018  Revised:March 15, 2019
中文关键词:粒子滤波  重采样  有效粒子数  残差  盲分离
英文关键词:particle filter  resampling  effective particle number  residual  blind separation
基金项目:中央高校基本科研业务费专项资金(2082604194194)
Author NameAffiliationE-mail
LIN Xiaomeng College of Electronics and Information Engineering, Sichuan University, Chengdu 610065, P. R. China x_meng_lin@163.com 
GAO Yong College of Electronics and Information Engineering, Sichuan University, Chengdu 610065, P. R. China gaoyong@scu.edu.cn 
Hits: 24
Download times: 11
中文摘要:
      在通信混合信号单通道盲分离技术中,粒子滤波盲分离算法是一大主流算法。针对粒子滤波算法必然存在的粒子退化问题,提出一种结合有效粒子数和残差重采样思想的改进重采样粒子滤波盲分离算法。所提算法在粒子退化,即有效粒子数减少至需进行重采样时,根据有效粒子数的值找出一定数量的权重大的粒子,按每个粒子的权重占总权重的比例复制一定数量的该粒子到新粒子集中。复制后将该粒子权重更新为剩余权重,把所有权重重新归一化,再用传统随机重采样方法复制少量粒子,完成重采样的改进算法,进而得到改进重采样的粒子滤波盲分离算法。仿真结果表明,在粒子滤波单通道盲分离应用中,提出的改进重采样算法与传统随机重采样算法相比,在复杂度近似相同的情况下误码性能提高了约1 dB。
英文摘要:
      In the single channel blind separation technique, particle filter algorithm was a mainstream method. In order to solve the problem of particle degradation, this paper proposed an improved algorithm which combined the effective particle number and residual resampling idea. When the number of effective particles reduced, resampling was invoked. This paper proposed that a certain number of particles with heavy weight were found according to the value of effective particle number during resampling process first. And these particles were copied in accordance with the ratio of each weight to the total weight and added to the new particle set. Second, these particles’ weights were updated to residual weight respectively and all weights were renormalized. Last, in order to complete the resampling algorithm, a small number of particles were copied. The simulation results showed that in the single-channel blind separation application, the resampling algorithm proposed had approximately 1 dB gain when complexity was the same as traditional random resampling algorithm.
HTML    View PDF   View/Add Comment  Download reader