Logo image
A Coefficient Ordering based Low Power Pipelined OFDM Transmitter Design for IEEE 802.16-2004 Physical Layer Baseband
Thesis

A Coefficient Ordering based Low Power Pipelined OFDM Transmitter Design for IEEE 802.16-2004 Physical Layer Baseband

Yang-Chun Fan
Masters, 國立清華大學, 通訊工程研究所
2004

Abstract

快速傅立葉轉換 正交分頻多工 傳輸機設計 OFDM FFT 802.16
Abstract IEEE 802.16 (wireless metro network, WirelessMAN or WIMAX called also) was designed for a metro network which can provide wide band access between several miles in the surface. The performance in data transmitting rate is almost the same as that cable, DSL, or T1 can provide, even can match that IEEE 802.11, wireless LAN system can provide. After the standard IEEE 802.16-2004 was released, the popular modulation technique named OFDM decrease the ISI substantially. According to this standard, I will introduce the transmitter design which fit to standard, inclusive of convolutional code encoder which can provide quite high coding gain, interleaver which can decrease the probability that the burst error may occur, and how to insert pilot tone into data subcarriers just like standard said, the baseband signal processing consists of Inverse Fast Fourier Transform and CP (cyclic prefix) insertion. Since the critical module among the transmitter system will be Inverse Fourier Transform, since IFFT acts just the same as FFT. I will introduce a novel coefficient ordering technique used pipelined FFT which can reduce the switching activity to achieve low power consumption, compare to former parallel FFT processor, the pipelined FFT provide real time property and data ordering technique can be implemented in this long point FFT. Since the only noise source in the transmitter is quantized noise, I will introduce the effect of finite wordlength and the comparison of SNqR in different length of word length also. Finally, after finishing the hardware description language (verilog) design, I will realize my design on the FPGA platform by using Quartus II tool, thus we can verify the full design in this platform. And others can modify or add new algorithm under this platform in the future.

Metrics

1 Record Views

Details

Logo image