CAJ | 학술논문

为实现对包括低密度奇偶校验( LDPC )码在内的一系列“稀疏几何编码”的识别，基于软解调序列，设计和实现了一种闭集集合应用背景下的识别方法。以校验约束关系作为切入点，首先回顾了基于硬解调序列的传统识别策略；接着分析和推导了软解调序列所应满足的校验约束形式；通过引入后验校验对数似然比( CLLR)这一统计量，最终设计了相应识别算法。在事先给定先验编码集合的闭集应用模式下，仿真结果显示，算法明显占优，识别增益在低信噪比环境下可达2~6 dB。
위실현대포괄저밀도기우교험( LDPC )마재내적일계렬“희소궤하편마”적식별，기우연해조서렬，설계화실현료일충폐집집합응용배경하적식별방법。이교험약속관계작위절입점，수선회고료기우경해조서렬적전통식별책략；접착분석화추도료연해조서렬소응만족적교험약속형식；통과인입후험교험대수사연비( CLLR)저일통계량，최종설계료상응식별산법。재사선급정선험편마집합적폐집응용모식하，방진결과현시，산법명현점우，식별증익재저신조비배경하가체2~6 dB。
To realize recognition of the sparse-graph coding including low density parity check ( LDPC ) codes,a finite set algorithm based on the soft-demodulation sequence is proposed. Through making coding parity-relationship as the key point, this paper first reviews the traditional recognition strategy based on hard-demodulation sequence. Then,it analyzes the coding check relationship which soft-demodulation se-quence should satisfy. Finally,it designs a new recognition algorithm by introducing a statistics index called posterior check-log-likelihood-ratio ( CLLR ) . The result of simulations which focalize around finite set shows that the new recognition algorithm based on soft-demodulation sequence is better than traditional method based on hard-demodulation sequence,with a recognition gain of 2~6 dB,particularly in low sig-nal-to-noise ratio( SNR) .