Design And Optimization Of Semiconductor Optical Amplifier Based Phase Regenerator

Explosive increasing of the data traffic on the Internet requires higher bit rate and larger capacity of the backbone network. For sake of the wide bandwidth, large communication capacity, low power loss, longer relay distance, better security feature and cheaper raw material, optical fiber communication system has been widely adopted as the solution of backbone transmission. Multi-Level phase modulation formats are widely used as the modulation format of the commercial backbone network due to their suppression of the dispersion effects in fiber. For the communication systems that use phase modulation formats, phase noise is the main element that restricts their capacity.Phase noise suppression and compensation techniques in optical fiber communication systems are called phase regeneration techniques. Nowadays there are a lot of phase regeneration techniques both in electric domain and optical domain. In electric domain, before detection, optical signals are firstly converted to the electric signals, and then the phase noise of the converted signals is compensated by the relevant coherent algorithms running on the DSP or FPGA board. In electric domain, the principle is mainly Optical-Electric-Optical conversion, while in optical domain, the principle is to realize the phase noise regeneration relying on the nonlinear effects of some optical devices. Although more ideal phase noise regeneration effect can be realized in electric domain, the upgrade of the bit rate of the optical fiber communication systems may be confined because of the electric bottleneck which is usually about40Gbit/s. However, optical fiber communication system is transparent to the bit rate of the signal, therefore the investigation of phase noise regeneration techniques in optical domain has been the focus in the academic world.In this thesis, the DPSK phase modulation format in high-speed optical fiber communication system and semiconductor optical amplifier devices are discussed. One phase noise regenerator structure whose basic principle is the cross gain modulation (XGM) effect of the SOA is focused.(This structure is called gain collision type phase noise regenerator in this thesis.) The working principle analysis, theoretical investigation and numerical simulation are conducted for this type of regenerator and the optimal design and working conditions are found out. The work in this thesis is very meaningful both in theory and in practical use because it can guide the use of this gain collision type of phase noise regenerator in the experiments.The following items are the main contents in this thesis (the items which are in bold are the innovative works).The background and value of phase noise regeneration techniques in optical fiber communication systems are fully discussed.The relevant modulation and demodulation techniques of DPSK signal are reviewed including the definition, feature, modulation and demodulation methods, Q factor and etc. The waveforms in time domain, spectrum in frequency domain and eye diagram of DPSK signal are obtained through simulation.Background knowledge relevant to this thesis is briefly discussed including phase noise principle, photon transition and radiation theory, structure, classification and amplification principle of semiconductor optical amplifier, nonlinear effects of semiconductor optical amplifier.Review the structure and principle of some type of all optical domain phase regenerator and show the advantages and disadvantages of them. Propose the SOA cross gain modulation effect based gain collision type of phase regenerator and make a comparison among them.The mathematic model of the gain collision type of phase regenerator is established. The optimal parameters of the SOA and the optimal working condition of this type of regenerator are found out in theory and I try to give the physical analysis of these optimal parameters.Some investigations which are relevant to SOA are also discussed. The topics consist of two parts:The first part is the design and weld of the driving circuit of the SOA device and the second part is about the theoretical investigation of the optimal conversion efficiency of SOA-FWM based all optical wavelength converters.