Research On Some Key Technologies Of All-optical Signal Process In Optical Communication System

People take the more and more high request to the communication bandwidth and signal processing speed with the development of internet and multimedia technology. The optical transmission networks as the foundation for internet and communication networks, are under pressure to carry massive amounts of data,which are developing towards high speed and large capacity. With the development of high-speed optical transmission technology, if the data switching at the core node still need to carry on the conversion of the O/E/O, the limited speed on signal processing must be a bottleneck to high-speed optical transmission. The key technologies of all-optical signal processing directly in the optical domain, which need not do O/E/O conversion, have been attracted wide attention. In this paper, all-optical multicasting and non-return-to-zero(NRZ) to return-to-zero(RZ) code conversion with multcasting techqiqes in all-optical signal processing are studied.When two optical fields having different wavelengths propagate simultaneously inside a fiber, the coupled nonlinear Schrodinger equations are applied to describe the cross-phase modulation(XPM) induced by coupling.The XPM spectrum of each pulse is expected broaden and develop a multipeak structure. Because the coupling nonlinear schrodinger equations are nonlinear partial differential equations, they do not generally lend itself to analytic solutions. Under the condition that the input signal is Gaussian signal, we use the common split-step Fourier method to solve the time dependence of the phase and XPM-induced frequency chirp, and then calculate the frequency interval between peak to peak. There is a big error between the calculation results and the simulation situation. In this paper, the causes of error produced are analyzed. The formula about the angular frequency of the largest ?rst-order gain is revised.16-Quadrature amplitude modulation(QAM) technology with higher spectral efficiency and better tolerance against nonlinearity has been an alternative to 100Gbit/s commercial systems. Therefor it has received much research. But in current conditions, It is lack of effective means for monitoring the 16-QAM signal,which can carry information on its pahse. In order to measure phase information of high-speed optical signal, we need to modulate the 16-QAM signals under existing experimental conditions. Hence, we carry out the research work on the 16-QAM modulation technology.The research work and results are summarized as follows:(1) For the first time, we study the impact of the pump and probe powers on the numbers of multicast channels, which is based on cross-phase modulation in a highly nonlinear ?ber. Wavelength multicasting is realized by appropriately controlling the powers of two beams. Our simulation work reveals that 10 multicast channels can be obtained with their Q factors being larger than six, if both pump and probe powers are properly selected. These wavelength channels of multicasting are appeared sequentially and positioned around the central wavelength of the probe on the blue-shifted and red-shifted sides. The pump and probe powers have obvious influence on establishing channels, but they have little influences on the central wavelength and the channel spacing.(2) Matching ITU grid wavelength multicasting based on cross-phase modulation is put forward for the first time. With a pump-modulated light and only a single continuous-wave probe we can achieve 10 multcasting channels, by fine adjustments of the pump and probe wavelengths, which cental wavelengths and channel spacings of each multicast channel from H35 to H44 match the C-band ITU grid. The channel spacing is approximately 0.8 nm when the probe and pump wavelengths are 1544.92 and 1554.78 nm, respectively. Through a wavelengths-swap, multicast channls can be converted toward the direction of long wavelength. So far, XPM in an optical fiber has been mainly used for wavelength conversion. However, the research on fiber XPM-based wavelength multicasting matching ITU grid in the C-band is not reported. The wavelength multicasting technique based on XPM is simpler and can offer more multicast channels than that based on four-wave mixing.(3) Every time we change the wavelength spacing of 0.4nm between probe and pump lights, and obtain 70 optical spectra, which indicate the wavelength differential between two beams may cause a cyclical optical spectrum evolution process of XPM. At present, there is no relevant report in the experimental research about the regular process of XPM. Simulation experiments prove that the frequency interval between peak to peak is the channel spacing. To calculate the channel spacing, we revise to the formula about the angular frequency of the largest ?rst-order gain that is derivated by G. P. Agrawal on assumption of vg1≈vg2. The revised formula and the simulation results are agreeable qualitatively.(4) NRZ-to-RZ code conversion with multicasting is investigated with a pump-modulated light and a single clock-wave probe. We obtain NRZ-to-RZ code conversion with 9 multicast channels and RZ signal pulse width is controlled by the clock pulse and output signal pulse width is adjustable. NRZ-to-RZ code channels have the same central wavelength and channel spcing as that the continuous-wave is a probe wave. 10Gbit/s and 20Gbit/s NRZ-to-RZ code conversion can be completed in the same fiber. The simulation results further prove that the NRZ-RZ code conversin and wavelength multicasting can be integreated together.(5) An initial investigation has been done on the theoretical of 16-QAM modulation and demodulation. The simulation model of 16-QAM transmitter/receiver are constructed using two Mach-Zehnder modulator by 4-ary electrical driving signals. The receiver use a 90 o optical hybrid and two balanced detection. The simulation results indicate that the constellation of 16-QAM in receiver is clear.