A novel quadrilateral companding transform for PAPR reduction in OFDM systems

One of the major limitations of Orthogonal Frequency Division Multiplexing (OFDM) signal is high Peak-to-Average-Power-Ratio (PAPR). Non-linear companding transforms that expand small signals and compress the peak signals reduce the PAPR significantly. The proposed companding scheme transforms the Rayleigh distributed probability distribution function (PDF) of original OFDM signal magnitude into quadrilateral shape PDF; hence the name “Quadrilateral Companding Transform (QCT)”. The proposed method provides better design flexibility to achieve an effective trade-off between PAPR reduction capability and bit-error rate (BER) performance. Theoretical performance of the proposed scheme is characterized by means of transformation gain, complementary cumulative distribution function (CCDF) of PAPR and companding attenuation distortion. The parameters of the proposed companding function are also calculated for optimal BER and PAPR performance. Simulation results confirm that the proposed scheme achieves better BER performance as compared to exponential companding transform, trapezoidal companding scheme and trapezium distribution based companding methods, while achieving almost same PAPR reduction capability. It is also shown that the proposed scheme is a generalized companding scheme and exponential companding transform, trapezoidal companding scheme and trapezium distribution based companding methods are the special cases of our proposed quadrilateral companding transform. BER performance of the proposed QCT is also analyzed over fading channel with and without including the effect of power amplifier.