Research On Recognition And Matching Of Variable-Rate Fiber Communication System

Large amount of information is produced every moment, in the development progress of modern society. The use of optical fiber communication technology makes the information flow up and realize its value. Meanwhile, the dataflows, which carry the increasingly diverse information, put forward higher requirements for transmission network on multi-service and intelligent aspect. The evolution of ptical transmission network, from early Plesiochronous Digital Hierarchy (PDH) to Synchronous Digital Hierarchy (SDH), and the later Multi-Service Transfer Platform (MSTP), is in order to meet these requirements.OFDM modulation has great advantages in optical transmission area, besides its wide range of application in high-speed wireless transmission. For instance, it has very high spectral efficiency, strong ability in resistance to intersymbol interference, and characteristics reducing the polarisation mode dispersion and the chromatic dispersion(CD). Therefore, the study in OFDM optical transmission technology in scenarios to achieve a multi-service channel the and smart self-adaptive transmission, do has important practical significance.This thesis describes the design ideas of a multi-channel and self-adaptive variable-rate optical OFDM transmission system, and study the variable-rate transmission system from two aspects-changing modulation formats and inserting redundant bits.In changing modulation formats, mathematical tools such cluster analysis and statistical moments computing are used to achieve the function of variable-rate; and in inserting redundant bits aspect, Parallel-XOR variavle-rate algorithm and curve deviation algorithm, which can reduce miscarriage of justice in non-ideal channels, are put forward. On this basis, we design system architecture, and the complete the simulation. Through the simulation, expected results were obtained.Finally, based on the results of optical OFDM variable-rate study above, several research directions are envisaged, such as Automation classifier and fine subcarrier variable-rate identification, which have practice value and room for improvement.