An approach based on trispectrum and rank statistics to testing departure from Gaussianity of stationary signals [radar signal processing]

• Published in: 2003 Proceedings of the International Conference on Radar (IEEE Cat. No.03EX695) pp. 133 - 137
• Main Authors: Zongchuang Liang, Xingzhao Liu, Yongtan Liu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2003
• Subjects: Algorithm design and analysis; Analytical models; Gaussian processes; Performance analysis; Radar signal processing; Signal processing; Signal processing algorithms; Statistical analysis; Statistical distributions; Testing
• ISBN: 9780780378704; 0780378709
• DOI: 10.1109/RADAR.2003.1278725

In this paper, a rank procedure based on trispectrum, to test the departure from the Gaussianity of stationary signals is presented. Both theoretical analysis and simulation results demonstrate that the proposed algorithm, to symmetrically distributed processes, has advantages in testing performance over the conventional approaches based on bispectrum, while maintaining the nominal level of significance, even for a relative small data size. Its results are better than those of the algorithms based on bispectrum and bootstrap.