Research and Design of Electronic Equalization Based on Most Likelihood Sequence Estimation
|School||Huazhong University of Science and Technology|
|Keywords||Electronic dispersion compensation Digital Signal Processing Balanced Maximum likelihood sequence estimation Channel estimation Fast Fourier Transform Viterbi algorithm|
Optical fiber communication systems because of its large capacity, a high rate of advantages has gradually become the most important part of the information transmission network . With the increase in the transmission speed , the low speed is not very serious dispersion problem is gradually revealed. Of 10Gb / s optical fiber and the device has been widely used , the full replacement of 40Gb / s or higher rates of fiber and the device is not economical. In order to improve equipment utilization , balancing technology gradually attracted people's attention . With respect to the optical signal equalizer , based on the electric field of the signal processing of the electrical domain equalization method is more economical easily achieved, but also a relatively high degree of automation . And balanced mainstream algorithm of the digital electric signal in the electrical domain equalization processing methods - the maximum likelihood sequence estimation, has been considered theoretically best balanced manner . Maximum likelihood sequence estimation the equalization method consists of two main modules : channel estimation module and Viterbi equalization module . Channel estimated module aims to channel to establish the mathematical model and simulation of digital signal on channel distortion process ; Witte than balancing module according to the channel estimated module to establish the mathematical model , the maximum likelihood estimate of the distortion sequence with the Viterbi algorithm restore the greatest degree of distortion data . This paper studies two different channel modeling model - probability model and finite impulse response filter model . Probability model, we use non-parametric modeling methods , modeling of different rate Fibre Channel ; finite impulse response filter model , this paper presents a channel estimation method based on a simplified fast Fourier transform . Secondly, this study achieved a conversion of the Viterbi algorithm , and two above channel estimation algorithm The Viterbi algorithm structures together to form two maximum likelihood sequence estimation equalizer . Through different simulation software to build an integrated simulation system to simulate these two equalizers and comparative analysis . Comprehensive simulation channel fast Fourier transform In this paper, based on a simplified the estimated equalizer outperforms the probabilistic model algorithm improved by an order of magnitude , and the computation time .