Performance improvement of OFDM system using CAZAC sequence and adaptive pre-distorter

OFDM (orthogonal frequency division multiplexing) communication system is very robust to the multiplex path interference, and its bandwidth efficiency is also very good for the design of the high speed communication system. However, high PAPR (peak average power ratio) is major drawbacks of OFDM system. In the nonlinear HPA (high power amplifier), the OFDM signal suffers serious nonlinear distortion and generally, this HPA is considered as memoryless device. So, combination of predistorter and CAZAC (Constant amplitude zero auto-correlation) matrix transform which has the autocorrelation characteristic and the regular power distribution both time domain and frequency domain are good solution to this high PAPR problem. But the situation will be different when the nonlinear HPA has memory property. In this paper, therefore, we analyze the PAPR reduction effects on Saleh HPA and memory polynomial HPA when CAZAC matrix transform and digital adaptive NLMS (normalized least mean square) predistorter are used at the same time. The memory polynomial model is used for the consideration of HPA memory effect. Also we show the performance variation in the HPA nonlinear characteristic according to change of IBO (input back off) value. From the simulation results for performance analysis, we can confirm that the proposed combined system shows better BER performance even in when the effect of HPA is considered.