作者: Michal P. Dlubek , Andrew J. Phillips , Eric C. Larkins
DOI: 10.1007/S11082-008-9242-5
关键词: Metric (mathematics) 、 Range (statistics) 、 Quality (physics) 、 Probability density function 、 Fisher information metric 、 Noise (electronics) 、 Mathematics 、 Stein's method 、 Applied mathematics 、 Bit error rate 、 Optics
摘要: A new and straightforward optical signal quality metric, which is based on the approximation of analytical probability density function by its Laguerre expansion, proposed. This expansion uses statistical moments to approximate true statistics, but, unlike well-known Gaussian approximation, higher order can be used as well. The proposed metric discussed compared with approximation. It shown that provides not only a good estimation bit error rate in range practical interest contrary correctly estimates optimum decision threshold accurately pictures shape noise distribution.