在对低密度校验码译码及密度演化算法研究的基础上,提出了分层译码的密度演化算法,对分层译码算法的噪声门限与收敛特性进行了仿真,并将分层译码与一般的置信传播及最小和算法做了性能比较.结果表明,分层译码算法具有快速译码收敛的优点,在不改变噪声门限的条件下,可使译码迭代次数减半.由于分层译码对降低译码器的复杂度起到了重要的作用,因此,文中提出的密度演化算法为实际应用提供了理论依据,能有效地估计低密度校验码分层译码的噪声门限和迭代次数.
A density evolution procedure for low-density parity-check (LDPC) codes under layered decoding was presented. The convergence threshold and characteristic of layered decoding were analyzed and compared with common belief propagation and min-sum algorithm. It is found that the layered decoding does not improve noise threshold but it can reduce the decoding iteration by half.