Deep Learning Based Stacked Sparse Autoencoder for PAPR Reduction in OFDM Systems

Orthogonal frequency division multiplexing is one of the efficient and flexible modulation techniques, which considered as central part many wired wireless standards. (OFDM) multiple-input multiple-output (MIMO) achieves maximum spectral efficiency data rates for mobile communication systems. Though it offers better quality services, high peak-to-average power ratio (PAPR) major issue that needs to be resolved in MIMO-OFDM system. Earlier studies have addressed PAPR OFDM system using clipping, coding, selected mapping, tone injection, peak windowing, etc. Recently, deep learning (DL) models exhibited improved performance on channel estimation, signal recognition, decoding, identification, end-to-end In this view, paper presents a new Hyperparameter Tuned Deep Learning based Stacked Sparse Autoencoder (HPT-SSAE) Reduction Technique The proposed model aims substantially reduce peaks signal. presented HPT-SSAE utilized adaptively create peak-canceling features input model, constellation mapping demapping symbols take place every individual subcarrier SSAE such way bit error rate (BER) systems are cooperatively diminished. Besides, enhance hyperparameter tuning process takes monarch butterfly optimization (MBO) algorithm. A comprehensive set simulations were performed highlight supremacy model. obtained experimental values showcased betterment over compared methods interms (BER), complementary cumulative distribution function (CCDF), execution time.