Nonlinear noise estimation and compensation in GFDM based communication systems for cognitive radio networks

Generalized Frequency Division Multiplexing (GFDM) might be a promising candidate for deployment in cognitive radio networks and for 5G wireless communication. It is a non-orthogonal multicarrier modulation scheme based on flexible pulse shaping. However, given by the high signal amplitude fluctuation, the performance of GFDM systems is highly degraded due to the distortion caused by the nonlinear nature of high power amplifier (HPA). In this article, we investigate the effects of nonlinearities in GFDM and introduce an iterative receiver scheme to reduce the error probability of GFDM signals undergoing nonlinear amplification. The proposed technique is based on successive estimation and cancellation of the in-band distortion introduced by the nonlinear nature of the HPA characteristics. The performance improvement of the iterative receiver structure is expressed by the bit error performance and error vector magnitude.