作者: Y.K. Kim , C.L. Weber
DOI: 10.1109/MILCOM.1989.104024
关键词:
摘要: A binary signal classification problem, where the received has different spectral correlations for hypotheses, is investigated. simple classifier, designated generalized single-cycle developed under additive white Gaussian noise (1) ideal, (2) unknown phase, and (3) phase epoch environments. As a specific example, classifier applied to SQPSK/2/sup k/PSK (k>or=1) modulation problem. The optimum choice of cycle relation between likelihood ratio are discussed. Emphasis placed on low signal-to-noise ratio. performance predicted by probability fixed rejection probability. >