Transformasi Wavelet dengan Teknik Clipping Filtering untuk Mereduksi PAPR pada OFDM

Abstract

Orthogonal Frequency Division Multiplexing (OFDM) is chosen as multiplexing techniques and broadly used in today’s radiocommunication environments to overcome spectrum insufficiency. With several superior advantages, however, OFDM is terribly affected by high peak to average power ratio (PAPR) due to offset frequency errors and local oscillator (LO) frequency synchronization errors. The high PAPR can cause nonlinear distortion, which outcomes in intermodulation and spectral leakage. This study aims to model the use of wavelet transform (discrete wavelet transform (DWT)) to replace Fourier transform (discrete Fourier transform (DFT)) that used in conventional OFDM, later in this paper is termed as DFT-OFDM. Clipping filtering techniques then applied to DWT-OFDM. The model was proposed to reduce PAPR in DFT-OFDM. The model was compared to DFT-OFDM using Matlab simulation method. The performance was evaluated using the Complementary Cumulative Distributive Function (CCDF) vs. PAPR. The results show that at PAPR 10-3for DFT-OFDM, it was produced PAPR of 10.6 dB whereas in DWT-OFDM, using Daubechies orde 7 (Daubechies7), Symlet orde 7 (Symlet7), Coiflet orde 2 (Coiflet2), were reached PAPR 4.8 dB, PAPR 3.3 dB, PAPR 3 dB, respectively. It means Coiflet2 providing the best PAPR reduction among other orthogonal wavelets. By applied clipping filtering to wavelet Coiflet2, it was produced PAPR of 2.9 dB for classical clipping and 2.8 dB for deep clipping. It show that wavelet Coiflet2 with deep clipping provided the best PAPR.