OPTIMAL QUANTIZATION OF BI-AWGN CHANNEL FOR TURBO DECODING
In this paper, we investigate the optimal quantizer design for turbo decoding with binary input additive white Gaussian noise (BI-AWGN) channel. The optimal design under MMSE (minimum mean square error) and MMI (maximum mutual information) criteria is derived and the results are compared with the simulation searched quantizers that have best frame error rate performance. Our results show that the MMI quantizer outperforms the MMSE one if the quantizer is used for turbo decoding. The performance of MMI quantizers is very close to the simulation searched quantizers (within 0.05dB) while the loss of MMSE quantizers is evident. To simplify the design, we also proposed a simple formula to approximate the MMI quantizer. The performance of this approximate quantizer is within 0.03dB from MMI quantizer. Moreover, we suggest that the quantization interval should be a slightly larger than the designed optimal value if the implementation error is a concern.
quantization minimum mean square error mutual information turbo code
Zhenqu Zhao Qijin Zhang Qiuping Huang Hongwen Yang
School of Telecommunication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China
国际会议
北京
英文
1-4
2008-09-26（万方平台首次上网日期，不代表论文的发表时间）