Abstract

Satellite communication channels are equipped with the Traveling Wave Tube (TWT) amplifiers which causes severe nonlinear distortions when working near saturation. A new predistorter is designed in this present study based on the double-mode adaptive technique. Firstly, a neural network is trained to design the amplitude predistorter, and the error-filtering scheme is applied to improve the classical structure of adaptive control, hence the convergent process becomes rapid using the LMS algorithm. Secondly, a real polynomial is applied to identify the phase character of TWT, and then the negative polynomial is copied to the transmitted path. Simulation results of 16-QAM modulations demonstrate high efficiency and low complexity of the proposed method, and the compensated system is characterized approximately by the ideal amplifier within the range of saturated input value.

Paper Details