10th EAI International Conference on Communications and Networking in China

Research Article

Method of Determining Number of Wideband Signals Based on Bootstrap Technique

  • @INPROCEEDINGS{10.4108/eai.15-8-2015.2261059,
        author={Jiaqi Zhen and Danyang Qin and Zhifang Wang},
        title={Method of Determining Number of Wideband Signals Based on Bootstrap Technique},
        proceedings={10th EAI International Conference on Communications and Networking in China},
        publisher={IEEE},
        proceedings_a={CHINACOM},
        year={2015},
        month={9},
        keywords={wideband signals bootstrap number of signals clustering},
        doi={10.4108/eai.15-8-2015.2261059}
    }
    
  • Jiaqi Zhen
    Danyang Qin
    Zhifang Wang
    Year: 2015
    Method of Determining Number of Wideband Signals Based on Bootstrap Technique
    CHINACOM
    IEEE
    DOI: 10.4108/eai.15-8-2015.2261059
Jiaqi Zhen1,*, Danyang Qin1, Zhifang Wang1
  • 1: Heilongjiang University
*Contact email: zhenjiaqi2011@163.com

Abstract

Traditional methods for determining number of signals usually need a lot of snapshots, or their performance will decrease seriously. In order to solve the problem, a method based on Bootstrap technique is proposed, the Discrete Fourier Transform is applied to divided the wideband signals into several nonoverlapping narrowband components, estimate the direction of an arbitrary signal, carry out eigen-decomposition of the covariance matrix and weight the eigenvectors by corresponding eigenvalues, subsequently employ the Bootstrap method to determining the number of signals of every frequency, integrate them to get the final result. The method is effective in the background of Gaussian white and non-Gaussian noise. Computer simulations prove the performance of the proposed method is better than the information theory criterion.