Near ML Detection of Nonlinearly Distorted OFDM Signals

Orthogonal frequency division multiplexing (OFDM) signaling exhibits high peak-to-average power ratio (PAR) which results in large sensitivity to nonlinear distortion created by the use of a high power amplifier (HPA) at the end of a wireless transmitter. In this work, in contrast with the popular approach of amplifier input signal clipping or low-PAR OFDM signal design, we utilize HPA characteristics to develop a low-complexity iterative detector of nonlinearly distorted OFDM signals that follows the principles of steepest descent search. Simulation studies indicate that the proposed method approaches maximum-likelihood performance and offers a significant gain over plain OFDM or other competing techniques.