Receiver based PAPR reduction in OFDM by compressed estimation of sparse signals

Abstract

This work establishes the design, analysis, and fine-tuning of a Peak-to-Average-Power-Ratio (PAPR) reducing system, based on compressed sensing at the receiver of a peak-reducing sparse clipper applied to an OFDM signal at the transmitter. By exploiting the sparsity of the OFDM signal in the time domain relative to a carefully defined clipping threshold, the method depends on partially observing the frequency content of extremely simple sparse clippers to recover the locations, magnitudes, and phases of the clipped locations of the PAPR reduced signal. We demonstrate that in the absence of optimization algorithms at the transmitter to confine the frequency support of clippers to a predefined set of reserved-tones, no other tone-reservation method can reliably recover the original OFDM signal with such low complexity. Emphasis is then given to designing different clipping signals that can embed a priori information regarding support, phase, and magnitude to the receiver, followed by more elaborate compressive sensing techniques for enhanced recovery. This includes weighted minimization for enhanced support recovery, phase-augmented-minimization for homogeneous clippers, and sparse lattice decoders for digital magnitude clippers. Finally, we adapt a greedy Bayesian compressive sensing technique to our system that exploits probabilistic information to enhance random sparse signal recovery from highly incomplete observations. Consequently, a Maximum A Posteriori (MAP) criterion is applied over a successively reduced search space for a practical balance between performance and complexity. We show that using such techniques for a typical OFDM signal of 256 subcarriers and 20% reserved tones, the PAPR can be reduced by approximately 9 dB at moderate Symbol Error Rate degradation and significantly lower complexity compared to standard tone-reservation based PAPR reduction systems.