CAJ | 학술논문

中短长度LDPC码采用BP迭代译码时,在低信噪比区,大量的经过多次迭代仍不能纠正的错误帧造成了平均迭代次数的大大增加;在中高信噪比区,比特对数似然比值的振荡造成了译码性能的降低.为了减少低信噪比区的平均迭代次数,提高中高信噪比区的译码性能,本文提出了一种LDPC码的信道自适应迭代译码算法.该算法采用基于校验和错误模式的预判决机制和消息加权平均算法,通过低信噪比区预判决机制的主导作用来减少平均迭代次数,通过中高信噪比区消息加权平均算法的主导作用来抑制因比特LLR值振荡而形成的错误帧,从而实现译码算法与信道变化特征的自适应,提高了译码效率.仿真结果显示,相对于BP译码算法,该译码算法在低信噪比区减少了平均迭代次数而译码性能没有显著变化,在中高信噪比区提高了译码性能而平均迭代次数没有明显增加.
중단장도LDPC마채용BP질대역마시,재저신조비구,대량적경과다차질대잉불능규정적착오정조성료평균질대차수적대대증가;재중고신조비구,비특대수사연비치적진탕조성료역마성능적강저.위료감소저신조비구적평균질대차수,제고중고신조비구적역마성능,본문제출료일충LDPC마적신도자괄응질대역마산법.해산법채용기우교험화착오모식적예판결궤제화소식가권평균산법,통과저신조비구예판결궤제적주도작용래감소평균질대차수,통과중고신조비구소식가권평균산법적주도작용래억제인비특LLR치진탕이형성적착오정,종이실현역마산법여신도변화특정적자괄응,제고료역마효솔.방진결과현시,상대우BP역마산법,해역마산법재저신조비구감소료평균질대차수이역마성능몰유현저변화,재중고신조비구제고료역마성능이평균질대차수몰유명현증가.
For short to moderate block-length LDPC codes, a large number of uncorrected error frames causes the increase of average number of iterations in the region of low SNR(Signal-to-Noise Ratio) and the oscillation of some variable nodes degrades the error performance in the region of medium to high SNR. In order to reduce the average number of iterations in the region of low SNR and improve error performance in the region of medium to high SNR, an iterative decoding algorithm with channel adaptation is proposed. The decoding method with pre-decision making based on error patterns of check-sum and weighted average message is adopted in this algorithm. Through the primary role of early decision method on reducing average iterations at low SNR and weighted average message on reducing error frames at medium to high SNR, iterative decoding with channel adaptation is realized and high efficient decoding is achieved. Simulation results show that, with the proposed algorithm, the average number of iterations is reduced in the region of low SNR without noticeable error performance degradation, and the error performance is improved in the region of medium to high SNR without obvious increase of the number of iterations compared with standard BP algorithm.