A modified ace method to reduce PAPR in OFDM- based systems

Abstract

Orthogonal Frequency Division Multiplexing (OFDM) technique appeared in history for a long time that has recently seen rising popular in wireless and wireline applications. An OFDM-based system provides greater immunity to multipath fading and impulse noise and eliminate the need of equalizer, while efficient hardware imple- mentation can be realized using Fast Fourier Transform techniques. However, one of major drawbacks of OFDM is the high fluctuation of the signal in time domain mea-

sured by Peak-to-Average Power Ratio (PAPR). Many approaches have proposed to deal with this problem and Active Constellation Extension (ACE) algorithm is one of the most effective methods. To reduce the PAPR, it adjusted intelligently the

constellation points into the increasing margin region without degrading the error probability for other data symbols. Although it increases the transmitted power, in practice, very large peaks occur rarely enough that these modifications have only a small impact on total transmitted power. In this thesis, we propose a modification of ACE in which the initial constellation points are shifted towards the margin region by some constant α greater than and closed to 1. We demonstrate the improvement

of the novel method by various experiments. We also provide a theoretical result that controls the effect of the signal clipping in ACE methods.