Research on BLMS Algorithm Based on Adaptive Predistorter

The characteristic of the radio frequency power amplifier varies with many kinds of factors,such as the environment temperature,power supply,etc.In order to guarantee the steady operation of the predistortion power amplifier,the adaptive performance of the predistortion system seems very important.In this paper,the predistorter is constructed based on memory polynomial.The indirect learning architecture is adopted to design the predistortion system.According to the non-line relation ship between the step size and the error,at the same time momentum factor is introduced we propose a new function of variable step size.The results shows this algorithm has a more perfect effect in improving the speed and Accuracy of the predistortion amplifier.