Compensation for Radio-over-Fibre Uplink Based on Hybrid Neural Networks

The radio-over-fibre （ROF） uplink, which combines the merit of optical fibre with that of microwave technology, can supply the high capacity of communication. However, there are two major issues： nonlinear distortion of the optical link and the multipath dispersion of the wireless channel, affecting the performance of the system. We propose an equalizer based on hybrid neural networks. The compensation needs no estimation of the channel. The simulated result shows that the ROF uplink can be adequately compensated and the performance of the equalizer depends on the channel noise.