Blind estimator of multiple optical signals with a detection matrix

Anhong Dang

2009-01-01

Abstract:An estimator model for multiple optical signals via a detection matrix is proposed and investigated. In the model, the intensity of the transmitted optical field is modulated by its own unique pseudo-random (PN) sequence and random binary data symbols, and then an estimated algorithm motivated by a signal subspace approach of using the sample matrix of the received signals is obtained under the hypothesis of a synchronous system. The optical channel is modeled using a log-normal distribution for the received signal intensity for low scintillation indexes. With the aid of numerical simulation, it is shown that the signal-to-noise ratio (SNR) can be estimated within 90% of the actual SNR after less than 18 frames. Moreover, it is shown that the method requires essentially no information about the propagation-channel characteristics and the amplitudes of signal and noise across the detector matrix.