Semi-blind compensation method for addressing memoryless nonlinearities

Nowadays, in order to guarantee both system performance and power efficiency, the resistance of nonlinear distortion produced by power amplifier (PA) has been a key issue in the wireless communication research. The traditional predistortion methods require prior knowledge of the amplitude, phase or bandwidth of the input signal, which is not very practical in the real world. To overcome it, we put forward a framework for compensating a nonlinear memoryless system in a semi-blind way. In this framework, by making use of the feedback branch, a nonlinear equation can be established to describe the input-output relations for the nonlinear system, and the gain of compensator can be iteratively obtained through solving this nonlinear equation with Newton method. Simulations are provided in order to verify the performance of this proposed framework and algorithm, where Saleh model is used as benchmark. Compared with traditional frameworks using the least mean square (LMS) algorithm similarly, our framework can achieve better performance in terms of mean square error (MSE) and latency without compromising the compensation effect.