Real-Time FPGA System for Modulation Format Identification of Optical Fiber Communication

Abstract

In flexible optical fiber communication systems, modulation format identification (MFI) is a useful feature. Although many MFI methods have been proposed, most of them are suitable for fixed channel conditions only. In addition, there exists no hardware-based implementation for identifying more than two formats. In this the sis, two methods of real-time identification are implemented and evaluated. Both of them passed software and hardware test. The method based on calculating informa tion entropy has the advantages of low hardware overhead and fast processing speed, and another method using artificial neural network (ANN) has the advantages of high accuracy and strong impairments tolerance. In the actual optical fiber commu nication system, different methods can be selected according to the characteristics of this system.