Probability density function fitting and fuzzy logic based coherent light communication system blind equalization method

Abstract

The invention discloses a probability density function (PDF) fitting and fuzzy logic based coherent light communication system blind equalization method, aiming at enabling a real part and a virtual part of the PDF at an output end of a light communication receiving terminal equalizer to match a transmitted known constellation. The method comprises the steps of estimating data PDF by adopting a Parzen window method, and adjusting a core size of the Parzen window by the fussy logic (FL). The method disclosed by the invention can obtain a faster rate of convergence and a smaller stable mean square error (MSE) than a traditional constant modulus algorithm (CMA) and a traditional multi-mode algorithm (MMA). The method is applicable to a polarization multiplexing (PolMux) orthogonal amplitude modulation (QAM) high-speed coherent light communication system.

Description

Technical field [0001] The invention belongs to the technical field of digital coherent optical communication, and specifically refers to a coherent optical communication system blind equalization method based on probability density function fitting and fuzzy logic. Background technique [0002] With the increasing demand for capacity of optical communication networks, many technologies have been proposed to improve transmission capacity and spectrum efficiency. For example, digital coherent receivers that incorporate digital signal processing (DSP) technology that can effectively eliminate propagation loss have completely changed the design of optical communication systems. At the same time, the polarization multiplexing (PolMux) coherent technology is widely used in coherent optical communication systems because it can double the system's spectrum efficiency. [0003] In addition, in order to meet the increasing capacity demand of optical communication systems, two technologies ...

AI Technical Summary

Problems solved by technology

[0005] However, since this technique cannot overcome the phase ambiguity problem, a carrier phase rotator is required to produce the correct constellation orientation

In addition, the key point of the Parzen window method is the selection of the window size, and the size of the optimal kernel is difficult to choose

Therefore, for different applications, different methods may be adopted, and adaptive adjustment and matching cannot be achieved

Images