抗差错算术码纠错性能分析  

Error correction performance analysis of error resilient arithmetic code

作　　者：王翾[1] 谢廷俊[1] 陆建华[1] 

机构地区：[1]清华大学电子工程系,北京100084

出　　处：《清华大学学报（自然科学版）》2007年第10期1661-1664,共4页Journal of Tsinghua University(Science and Technology)

基　　金：国家自然科学基金资助项目(60328103;60532070)

摘　　要：为了应用抗差错算术码进行系统设计和确定信源信道联合编码方案,需要较准确估计抗差错算术码的纠错性能,但目前,抗差错算术码的纠错性能只有仿真结果可以参考。该文以堆栈算法为例,通过理论分析得到了抗差错算术码序列差错概率上界,该上界与(1-R)成反比(R是抗差错算术码的编码效率)。同时,分析过程还为解码过程参数设置提供了理论依据。仿真结果证明了该上界是对算术码纠错能力的良好近似,可以用于编码方案的优化设计。Error correction performance analysis is essential for optimizing parameters in designs of joint source channel coding schemes. However, in most joint source channel coding schemes, the relationship between the error correction performance and the encoding rate can only be assessed through empirical tests. This paper presents a theoretical analysis of an error resilient arithmetic stack decoding algorithm as an example. The upper bound on the sequence error probability is shown to be inverseiy proportional to （1-R）, where R is the error resilient code coding rate. Simulations verify the accuracy of this upper bound. The analysis method also provides a theoretical basis for parameter optimization of joint source channel coding scheme designs.

关 键 词：信源编码理论 抗差错算术编码 序列译码 序列差错概率 

分 类 号：TN911.21[电子电信—通信与信息系统]