Fiber-optic communication systems is playing an important role in digital communication networks. The trend is heading for higher bit rate and longer distance. In order to achieve this goal, systems need to prevent the degradations from many impairments, including noise, dispersion, fiber nonlinearities, etc. Thus, the monitoring and compensation for these effects are very important for the system. However, these impairments are not static because for reconfigurable networks, the optical link itself is not static. Moreover, the high bit rate system is very sensitive to even little variations which were regarded as negligible in lower bit rate systems. Hence, an inline monitoring method is crucial for future dynamic compensation....[ Read more ]

Fiber-optic communication systems is playing an important role in digital communication networks. The trend is heading for higher bit rate and longer distance. In order to achieve this goal, systems need to prevent the degradations from many impairments, including noise, dispersion, fiber nonlinearities, etc. Thus, the monitoring and compensation for these effects are very important for the system. However, these impairments are not static because for reconfigurable networks, the optical link itself is not static. Moreover, the high bit rate system is very sensitive to even little variations which were regarded as negligible in lower bit rate systems. Hence, an inline monitoring method is crucial for future dynamic compensation.

Asynchronous amplitude histogram has been attracting attention because it does not require the expensive cost on high-speed circuits for clock recovery. Based on asynchronous histograms, average Q factor can be obtained for bit-error-rate (BER) performance estimation. However, average Q factor is sensitive to dispersion, and fails to provide direct measurement on the dispersion. Thus, its application on the performance monitoring is limited.

In this thesis, we proposed that the cross-point count in an asynchronous amplitude histogram is strongly related to the pulse rise time, which is sensitive to the dispersion effect. We show that as the pulse is broadened by dispersion, the pulse rise time and the corresponding cross-point count will be increased. We develop the an analytical method to measure the rise time to bit period ratio by using the cross-point count ratio for both non-return-to-zero (NRZ) format and return-to-zero (RZ) format. We also propose a noise-correction method to retrieve the cross-point count ratio that is only contributed by the pulse rise time. Our method is also bit rate independent and protocol transparent, thus suitable for future all-optical networks ( AONs) .

The proposed method is demonstrated with computer aided design tools for optical communication systems. Proof-of-principle experimental results support the feasibility of the measurement and correction method.