Journal of Systems Engineering and Electronics ›› 2011, Vol. 33 ›› Issue (9): 2117-2121.doi: 10.3969/j.issn.1001-506X.2011.09.37

• 通信与网络 • 上一篇    下一篇

基于二阶循环统计量的BPSK参数估计

于宁宇1,2, 马红光1, 石荣2, 石磊3
  

  1. 1. 第二炮兵工程学院101教研室, 陕西 西安 710025
    2. 中国电子科技集团公司信息综合控制国家重点实验室, 四川 成都 610036
    3. 中国人民解放军96656部队, 北京 100028
  • 出版日期:2011-09-17 发布日期:2010-01-03

Parameter estimation of BPSK via secondorder cyclic statistics

YU Ning-yu1,2, MA Hong-guang1, SHI Rong2, SHI Lei3
  

  1. 1. 101 Section, The Second Artillery Engineering College, Xi’an 710025, China
    2. National Information Control Laboratory, China Electronics Technology Group Corporation, Chengdu 610036, China
    3. Unit 96656 of the PLA, Beijing 100028, China
  • Online:2011-09-17 Published:2010-01-03

摘要:

针对复值二进制相移键控(binary phase shift keying, BPSK)信号,推导了其无共轭项和带共轭项的二阶循环累积量表达式,分别研究了时延对其循环频率特性的影响,得出了独立的载频和码片时宽估计算法,即可以利用无共轭项二阶循环累积量时延零切片估计载频,可以利用带共轭项二阶循环累积量时延非零切片估计码片时宽,并推导了最优时延的选取,进而提出了一种更优良的码片时宽估计算法。仿真实验表明,提出的载频估计算法与循环谱算法比较,虽有2 dB的损耗,但是计算复杂度减小|码片时宽估计算法性能比利用循环谱的估计算法提高5 dB以上。

Abstract:

The expressions of non-conjugated and conjugated secondorder cyclic cumulants are deduced based on complexvalued binary phase shift keying (BPSK) signals. The influence of different timedelays on the characteristics of cyclic frequency is studied, and the independent algorithms are proposed to estimate carrier frequency and chip width. In other words, carrier frequency can be estimated based on nonconjugated secondorder cyclic cumulants with a timedelay to be zero|while chip width can be estimated based on conjugated  second order cyclic cumulants with a timedelay to be nonzero. To improve the performance, the optimum timedelay is deduced, and then a much better chip width estimation algorithm is proposed. Simulations verify that though the performance of the proposed carrierfrequency algorithm has 2 dB loss, computational complexity is much simpler than the cyclicspectrum algorithm, and the performance of the proposed chip width algorithm has a more than 5 dB increase compared with the cyclicspectrum algorithm.