A Direct Learning Structure for Adaptive Polynomial-Based Predistortion for Power Amplifier Linearization

A new polynomial predistortion technique for linearizing nonlinear power amplifier is introduced. The proposed predistorter is based on direct learning, whereas most other techniques are based on the indirect learning architecture. The resulting structure of the proposed method is simpler than the existing ones. Computer simulation shows that the proposed method is robust to the initial conditions and attains comparable performance to other methods based on the indirect learning architecture under normal conditions.