A Novel Receive Algorithm for Clipped OFDM Signals

A major drawback of the Orthogonal FrequencyDivision Multiplex (OFDM) principle is the high dynamic range of the transmit signal. The transmit Power Amplifier (PA) and other elements in the signal path have to be operated at high Input Power Backoff (IBO), which decreases the efficiency of the PA and leads to higher costs. This paper presents a novel receive algorithm for OFDM signals with limited dynamic range. Similar to existing estimators for clipped real processes, a closed-form expression of a Bayesian estimator for more practically relevant clipped complex Gaussian processes, observed over the Additive White Gaussian Noise (AWGN) channel, is derived. The estimator is based on the distribution of the real and imaginary parts of the complex baseband OFDM signal. It recovers the original undistorted process by minimizing the mean-square estimation error. It is shown that this estimator significantly improves the error rate of OFDM systems impaired by clipping at the transmitter.