BASEBAND ADAPTIVE PREDISTORTION FOR MC-CDMA SIGNALS

摘要

Predistortion is a signal processing technique used to reduce the nonlinear effects of the transmitter. This paper presents simulation results of several adaptive baseband signal predistortion methods for MC-CDMA signals. Our results show that, contrary to popular belief, lookup table based methods can converge faster and perform well compared to Volterra series based methods. We test Park and Power's parallel Volterra series pre-distorter (which has the highest convergence speed in past study), Jeon's fixed point lookup table predistorter, and FLUT lookup table predistorter. FLUT lookup table pre-distorter requires much less training and computation time compared to all other methods, including Park and Power's parallel Volterra series predistorter trained with RLS method. All predistortion methods we tested achieve the same near ideal predistortion results in terms of minimum total degradation(MTD), reducing the minimum total degradation by nearly 5dB, showing that predistortion is an effective way to improve the quality of MC-CDMA signals.