Blind estimation and mitigation of nonlinear channels

Nonlinear distortions in the analog frontend are becoming a growing problem. With rising requirements it becomes more expensive to provide components with sufficient linearity. Especially in the receiver, where out-of-band radiation is not an issue, relaxing the requirements on the analog components could significantly reduce development and production costs as well as increase the energy efficiency. There are numerous algorithms to mitigate the effects caused by nonlinearities. This paper presents a method for the blind estimation of nonlinearity parameters which can be applied to a large class of nonlinearities. The estimator is integrated in an OFDM system with nonlinearity mitigation and it shows that there is negligible performance degradation compared to perfect knowledge. The Cramer-Rao lower bound is derived and it is shown that the estimator reaches it in the high SNR region. Special care was taken to provide a solution with low computational complexity that can be readily implemented on current platforms.