Optical Fiber Dispersion Monitoring

The rapid growth of Internet and wireless applications has resulted in the current network infrastructure to quickly approach its limits. The continually increasing demand for higher bandwidth is directly related to an increase in bit rate across the fiber network infrastructure. The increase in bit rate makes results in these systems being very sensitive to optical fiber degradation\u27s such as chromatic dispersion and nonlinearities. Minimizing these degradation\u27s poses a difficult challenge, due to these degradations not being constant, but varying with time due to changing parameters such as change in transmission path distance, temperature change, and change due to maintenance or aging of the fiber network. A viable solution to minimizing the effect of these degradations is to monitor the signal quality in the system and dynamically compensate for the variations in the degradations. By utilizing the fact that the low and high bits of a received bit stream are affected differently during transmission through an optical fiber, a technique was developed that can predict the amount of chromatic dispersion present in a fiber link. Automating an optical fiber network provides a performance benefit across the network, making Optical fiber monitoring research a very interesting topic. However, monitoring an optical fiber channel is not easy as the transmitted data is subject to both chromatic dispersion and nonlinearities, as well as the interaction between the two. The impact of these degradations increases as the length of the fiber increases. In this research project, we created a monitoring technique in which the relative distributions of the high and low bits in a 10.7Gbps binary NRZ signal are analyzed. A model was then created that uses the variance of the distributions to estimate chromatic dispersion, with a margin of error of 3% and a monitoring sensitivity of 900ps/nm independent of transmission distance