Channel Identification of communication system with nonlinear power amplifier

In communication system, power amplifiers (PAs) are important components and inherently nonlinear. In such systems, when a nonconstant modulus signal goes through the nonlinear PA, spectral regrowth (broadening) appears in the output. For channel identification which is based on the observation of system output, the spectral information contained in the outband would help to improve the estimation accuracy. Given that training input symbols are available, we introduce two channel identification algorithms by exploiting outband radiation for two cases. Specifically, when the PA input/output relation is fully known, our channel identification is based on the application of best linear unbiased estimators (BLUE); for a more common case, when the PA input/output relation is partially known, we propose an iterative channel identification algorithm. Our simulation results demonstrate the performance of our new approaches.