Digital Signal Processing Techniques And Its Simulations For Coherent Optical Transmission System

Coherent optical transmission system has captured intensive attention for its higher spectral efficiency and greater bit rate. But a serial of new challenges just line up, like nonlinearity induced by high in-fiber power, chromatic dispersion (CD), frequency offset (FO), laser-linewidth-induced phase noise, etc. While the impact of CD, for instance, could be cancelled by optical devices or modules, it is more than difficult to cope with the influence of nonlinearity and/or frequency offset with potential optical devices. However, along with the fast growing digital signal processing (DSP), DSP related techniques set solid underlying foundations for coherent detection, shaping the next-generation optical transmission system. Thus, this paper focuses mainly on four major problems in optical transmission, i.e., nonlinearity, chromatic dispersion, frequency offset and phase noise.The paper begins with the introductory chapter presenting the basic knowledge and conventions necessitated. Following that come the compensation algorithms based on DSP, expected to mitigate the effects of the four challenging problems in coherent optical transmission. Admittedly, of all algorithms and schemes discussed, some show less practicality because of either high complexity or poor performance. Hopefully, this leaves some room for the author to make some improvement and modifications to some specific algorithms. The primary modification is made to the backward propagation (BP) used to compensate the nonlinearity. Specifically, the conventional scheme requires several iterations within one span to reduce the nonlinear phase noise(NLPN), whereas one operation is needed within such a span, removing the nonlinear phase noise almost perfectly. What highlights the paper is that a frequency offset estimator (FOE) is proposed by the author to estimate the frequency difference between the transmitter and the local oscillator (LO) involved in QAM modulated signal of arbitrary level. The proposed FOE bases its operation on fast Fourier transform (FFT), effectively lessening the FO by introducing two steps into this algorithm, i.e., a coarse step and a fine step. In comparison to other FOE for QAM, this proposed scheme exhibits fix latency, lower requirement for logic units and stable performance, etc. The third aspect is that the paper deals with the improvement to FOE and phase estimator (PE) for PSK modulated format. Deductive derivations and numerical simulations have jointly confirmed that when lookup table (LUT) based FOE and PE are employed, the amplified spontaneous emission (ASE) noise can only degrade the phase component of samples instead of both phase and amplitude when fourth-power used. Therefore, the LUT based algorithms exhibit higher accuracy and thus require lower amount of logical units.To evaluate the overall performance of the algorithms discussed in the paper, simulation of a16-QAM transmission system with the four corresponding attributes is established. It is numerically and graphically shown that the algorithms could retrieve and recover the bit sequence transmitted with the bit error rate of2.1×10-3.