Comparative Analysis of PAPR Reduction Method using Windowing Technique and Digital Predistortion Technique for OFDM Systems

Digital transmission known as orthogonal frequency division multiplexing (OFDM) is primarily employed in wireless communication because of its capacity for data transfer and resistance to multipath fading. The OFDM signal’s primary constraint is its relatively high Peak to Average Power Ratio (PAPR). The peak window, with its flexible interoperability and easy to implement, is an effective measure of PAPR reduction technique. When employing SLM (Selective Mapping) to reduce PAPR, researchers have utilized a significant amount of side information (SI). When reducing PAPR utilizing a partial transmit sequence (PTS), the researcher considered an average PAPR decrease at the expense of thoroughly looking through every rotation phase combination. This paper proposes the use of Peak window functions for PAPR reduction. When a high-power amplifier (HPA) receives an OFDM signal with a high PAPR, it will cause highly nonlinear distortions that will raise the bit error rate (BER) of the entire system thereby reducing HPA power efficiency. In order to produce an OFDM signal with lower PAPR and improved BER, PAPR optimization with error vector magnitude (EVM) constraints has been investigated in this study. In order to counteract the effects of nonlinearities in a power amplifier (PA), the research suggests designing, training, and testing a neural network to apply digital predistortion (DPD). The neural network-based DPD (NN-DPD) offline training is the main topic of the paper.