New Predistortion Method for RF Power Amplifier Linearization

This paper proves that the sequence of two memoryless nonlinear systems, predistorter and RF power amplifier, is commutative through theoretical derivation. For the disadvantages of frequently used predistortion structures, this paper proposes a new predistortion structure combining direct learning structure with indirect learning structure, which is simple and easy to implement. The RF power amplifier predistortion linearization method is extended to the situation with memory effect. Simulation results show the correctness of this predistortion linearization method. Its linearization performance is better than that using the direct inverse indirect learning structure.