Over the past few decades, research has been performed on the use of soliton signals in optical communication systems. This potentially allows an optical signal to propagate along a fiber line with minimal need for signal restoration through electrical conversion. Various sources of noise can cause a soliton signal to become distorted and degrade as it propagates however, so solutions have been investigated to attenuate and separate the noise generated. Since one of the dominant sources of noise generated is through the amplified spontaneous emissions (ASE) of optical amplifiers currently in use, solutions were explored that could potentially involve a variety of devices. These solutions have shown that the noise as well as the resulting bit error rate (BER) can be improved to acceptable levels far beyond that of non-filtered systems.;An approach implementing a modification to the sliding-frequency filter method has demonstrated a great potential towards separating the noise components in a soliton transmission system. Previous experiments included the use of an acousto-optic modulator to provide the necessary frequency shift, followed by a fixed-frequency band-pass filter. After several considerations, it is determined that acousto-optic tunable filters (AOTF)s are capable of providing the narrowband filtering capabilities required for the design as well as the necessary frequency shifting aspects of a modulator. To this end, a variety of materials applicable for use in non-collinear applications are taken into consideration. MATLAB has been used to provide a means to compare the differences between several acousto-optic materials at once.;In order to explore the potential improvements of this filtering approach, simulations exploring the effectiveness of the filter design at a data rate of 10GB/s were performed using OptSim with parameters derived around TeO2 as an AOTF material base. To realize a proper noise model in the simulation, the amplified spontaneous emission noise was constructed using a full-pump model structure for each amplifier. As a result, the simulated signal was maintained at an acceptable bit error rate of 10-9 or lower for 2600km without the use of signal recovery and regeneration components. The simulated performance of this filter approach exhibits traits that are comparable to static band-pass filtering techniques as well as basic sliding-frequency filter approaches.
展开▼
