On Companding Schemes for PAPR Reduction in OFDM Systems Employing Higher Order QAM

Large peak-to-average power ratio (PAPR) of the signal is an imperative problem in orthogonal frequency division multiplexing (OFDM) systems, as it impedes the efficient design of their analog front end. Companding is a well-known technique that reduces the PAPR by using a deterministic amplitude transform, which is designed by assuming Rayleigh distribution for the signal amplitude. But in OFDM systems employing higher order QAM, like 16- or 64-QAM, this probabilistic model is insufficient to model the true characteristics of the signal. This is due to the stochastic nature of average symbol power, which alters the complete amplitude distribution from one OFDM symbol to another. This leads to degradation in the companders' PAPR reduction performance. In this paper, we propose two novel schemes that modify the companding operation to accommodate the randomness of symbol power. Probabilistic analysis of symbol's average power, in relationship with the amplitude of its constituent samples, is carried out. This yields the theoretical framework employed in the design of the proposed low-complexity solutions. The proposed schemes essentially make the companding operation itself nondeterministic/adaptive, i.e., it adjusts to comply with the changing symbol amplitude distribution during application runtime.