Neuro-Interpolator for Holographic Data Storage

The neuro-interpolator (NI) integrating the interpolator with neural network equalizer is proposed to reduce the influence of interpixel interference in holographic data storage. The simplification of NI is performed by using a hybrid genetic algorithm based on a new mutation method. The bit error rate performance of the simplified NI is superior to that of the interpolator alone. Furthermore, the number of connections in the simplified NI is reduced to about half of that obtained by using a conventional mutation method.