SCR-Based Tone Reservation Schemes With Fast Convergence for PAPR Reduction in OFDM System

The signal to clipping noise ratio (SCR)-based tone reservation scheme is a promising candidate for peak-to-average ratio (PAPR) reduction of orthogonal frequency division multiplexing signals, but it suffers from low convergence rate caused by the nonoptimal scalar scaling factor. In this letter, we first derive a scaling SCR (S-SCR) scheme, in which the scaling factor is an optimized vector with peak regeneration constraints. To further improve convergence rate and eliminate multiple peaks in each iteration, we then propose the multiple scaling SCR (MS-SCR) scheme with an augmented scaling factor in matrix form. Numerical results demonstrate the superiority of proposed schemes compared with the conventional ones, including better PAPR reduction performance with fewer iterations and comparable bit error ratio performance. Furthermore, the computational complexity reduction ratio of the proposed S-SCR and MS-SCR schemes are increased by 38.52% and 27.36%, respectively.

I. Introduction

Orthogonal frequency division multiplexing (OFDM) signals have been extensively adopted in wireless communication systems. However, OFDM signals often suffer from high peak-to-average ratio (PAPR) which leads to significant in-band distortion and out-of-band (OOB) radiation. Consequently, various PAPR reduction techniques have been proposed, such as clipping [1], coding [2], companding [3], selective mapping (SLM) [4], and partial transmit sequence (PTS) [5].