Clipping Noise Reduction for OFDM by Hybrid-Symbol-Reconstruction

• Main Authors: Li Cho, 卓立
• Other Authors: Chau-Yun Hsu
• Format: Others
• Language: en_US
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/95747441064217225888

碩士 === 大同大學 === 通訊工程研究所 === 98 === Clipping is a simple and efficient solution to orthogonal frequency division multiplexing (OFDM) systems suffered from high Peak-to-Average Power Ratio (PAPR). However, clipping induces a noise of in-band distortion and out-of-band radiation that degrades the system performance. In this thesis, we proposed a novel application of Papoulis-Gerchberg algorithm (PGA) for reducing the clipping noise, and improved it with decision-aided reconstruction (DAR) and useful specification characteristics to attain hybrid-symbol-reconstruction (HSR) ultimately. The proposed HSR is applied to IEEE 802.16d-2004 scenario as an example. In our clipping model (clipping ratio is 4 dB,) the simulation results show the performance can be recovered to within 5 dB of the non-clipped case without additional computation complexity relative to conventional PGA.