A review of intelligent predistortion methods for the linearization of RF power amplifiers

Wireless communications need linear and efficient RF power amplifiers. With the emergence of new generations of wireless devices and modulations that have non-constant envelopes, more attention has been given to the importance of the linear behavior of power amplifiers. Various approaches exist for the linearization of RF power amplifiers. However, the desirable method among these is the one which can be applied intelligently in order to reduce system monitoring and human error in the operation of the system. Of all the linearization methods, the predistortion technique is the one which can be applied digitally, and which can be implemented and executed by means of intelligent tools such as the Genetic Algorithm, Neural Networks and the Neuro-Fuzzy systems. By applying each of the aforementioned tools, an effective step is taken forward in making the wireless communication systems intelligent. In the same regard, it is hoped that all the sectors of wireless communication systems can be made intelligent so that the need for human monitoring of these systems can be totally eliminated. In this article, the principles related to the implementation of the predistortion method by means of intelligent tools will be investigated and the improvement in linear behavior by using each of these tools will be evaluated through some examples.