Reduction of PAPR Without Side Information for SFBC MIMO-OFDM Systems

A novel peak-to-average power ratio (PAPR) reduction scheme designated as extended selected mapping (eSLM) is proposed for space-frequency block coding (SFBC) multi-input multi-output orthogonal frequency division multiplexing systems. In the eSLM method, extension matrices comprising amplitude extensions and phase rotations are constructed to indicate the selected signal index without the need for side information and to minimize the PAPR, respectively. To reduce the computational complexity incurred by the inverse discrete Fourier transform operation in generating the candidate signals, a low-complexity eSLM scheme (LC-eSLM) is developed by constructing equivalent candidate signals in the time domain. Notably, the extension matrices in both schemes preserve the orthogonality of the SFBC code, thereby facilitating low-complexity decoding. The simulation results show that the proposed eSLM scheme not only outperforms existing blind SLM-based methods. Compared with the costly ordinary SLM scheme, the eSLM scheme has a lower computational complexity with a performance loss of less than 0.3 dB and requires no side information. Furthermore, the computational complexity of the LC-eSLM scheme is around 40%-50% lower than that of the eSLM scheme with only a marginal degradation in the PAPR reduction performance.

I. Introduction

Orthogonal frequency division multiplexing (OFDM) has been widely adopted as an attractive technique in modern wireless communication due to its robustness against frequency selective fading channels and its potential for achieving high data rate [1]. However, OFDM transmission systems suffer a high peak-to-average power ratio (PAPR), which results in severe in-band distortion in the nonlinear region of the power amplifier [2]. The literature contains various PAPR reduction methods for overcoming this problem, including clipping [3], selected mapping (SLM) [4], companding [5] and partial transmit sequences (PTS) [6].