The analytical expression for probability distribution density of transmission coefficient for passive linear channel of fiber-optic systems functioning under the influence of mechanical load has been obtained in the research. It has been shown that probability distribution density is sensitive to the amount of fiber microcracks and the amount of fading on each of them, and also includes the parameters that are determined by the optical fiber drawing technology.
The expression for electron count optimal threshold value has been obtained in the article, the expression for evaluation of potentially achievable average error probability value during binary signals differentiation by an optical receiver with consideration of synchronization error has been refined. It has been shown that a time interval error switches the receiver into a mode that does not provide minimal average error probability. This explains a significant increase in error probability, which is mostly caused by errors during transmissions of “zeros” due to a lowered count threshold value. The obtained results enable us to determine synchronization error exchange conditions on optical signal power with a predetermined average error probability value, and to substantiate synchronization precision tolerances.
Proc. SPIE. 10774, Optical Technologies in Telecommunications 2017
KEYWORDS: Signal to noise ratio, Photodetectors, Statistical analysis, Error analysis, Interference (communication), Receivers, Telecommunications, Signal processing, Nonlinear optics, Signal detection
The problems of statistical analysis of a typical receive optical module (TROM) in the detection of a random optical signal and the synthesis of an optimal algorithm for discrimination of the optical signals in a binary communication system are solved in a paper. The envelope of optical field intensity is described by the Nakagami distribution density. Conditional Poisson distribution of the photoelectron counts and its asymptotic approximation, during limitation of the detection process by thermal noise is obtained. Statistical model of the signal and noise mixture at the TROM output consider the dynamic range of optical signal. The synthesis of the Bayesian algorithm for signals discrimination is performed taking into account the restrictions imposed by the preset TROM structure. The average error probability of signals discrimination is estimated. Conditions for feasibility of the synthesized algorithm are discussed.