Peak Power Reduction of OFDM Systems Through Tone Injection via Parametric Minimum Cross-Entropy Method

It is well known that the major drawback of orthogonal frequency-division multiplexing (OFDM) is its high peak-to-mean envelope power ratio (PMEPR). One of the promising schemes proposed for alleviating the high PMEPR of OFDM signals is tone injection (TI). In this paper, we propose a novel low-complexity algorithm for PMEPR reduction using TI. This algorithm first formulates the TI scheme as a combinatorial optimization problem and then finds a near-optimal solution to this problem via parametric minimum cross-entropy (PMCE) method. We also demonstrate that by effectively exploiting TI schemes in conjunction with cross constellations, it is possible to reduce the peak power of the system by more than 2 dB compared with sign selection schemes over conventional quadrature-amplitude modulation (QAM) constellations. Simulation results show that our algorithm outperforms similar existing algorithms, including the cross-entropy (CE) TI algorithm, the greedy sign selection algorithm, and the derandomization sign selection algorithm, in terms of PMEPR. We also present a threshold-based PMCE algorithm for TI that results in significant complexity reduction. Numerical results show that to achieve a prescribed PMEPR threshold, this threshold-based PMCE algorithm requires less computational complexity than the CE TI algorithm.