提出了一种新的联合译码方案。该方案结合了卷积码和Turbo码译码算法的优势，在译码端采用三个译码器，前两个译码依然为经典Turbo码译码结构并采用Log-Map算法，最后一个译码器接收译好的信息位、校验位和外信息值，采用维特比译码算法。经过迭代译码，能够进一步提高传统Turbo译码的性能。仿真结果证明在10次迭代以前，至少能够获得0.1dB的编码增益。

The article proposes a novel combined decoding scheme. This structure combines the superiority of concolutional code and Turbo code. There are three decoders. The first two components are arranged as the classical parallel concatenated decoding scheme and using the Log Maximum A Posteriori ( Log -Map) algorithm. The third receiver who is applied with soft output Viterbi algorithm ( SOVA) takes over the system bit,the first parity bit and priori information from the first two decoders. It can improve conven-tional Turbo code performance further. Simulation result shows that it achieves 0. 1dB performance gains at least before 10 iterations.

Turbo码的优越性使其在3G、4G及卫星通信等领域应用广泛。文章在分析了几种算法的基础上对Max-log-MAP算法进行了改进,引入迭代译码停止准则,并对算法进行了定点和加滑动窗处理,从而降低了译码复杂度,减少了译码时间,使Turbo译码容易通过硬件实现。

Turbo code is widely used in the third and fourth generation mobile communication and satellite communication systems because of its excellent per-formance. The thesis has improved the max-log-MAP algorithm after analyzed several decoding algorithms. The thesis has adopted stop criterion, fixed-point data and sliding-window schemes to reduce calculation, storage and delay of decoding. So Turbo code is im-plemented easily in hardware.

Turbo码具有接近香农极限的译码性能。由于这种优异特性,空间数据系统咨询委员会(CCSDS)在1999年就把Turbo码加入了遥测信道编码建议书。介绍了CCSDS标准的Turbo码编码结构和译码原理,研究了影响其性能的各个重要参数,如信息块长度、迭代次数和码率等,通过Matlab软件对不同参数条件下的译码性能进行了仿真,对仿真结果进行了详细分析,给出了深空通信系统中参数选择的建议。

Turbo code has a very good performance which is close to the Shannon limit.In 1999,CCSDS(Consultative Committee for Space Data systems) has added Turbo code to the recommendation of Telemetry channel coding due to its excellent properties. This paper introduces the coding structure and decoding principle of Turbo codes of CCSDS standard,and studies important parameters which impact its performance,such as block length,iteration number,bit rate,etc.The decoding performances under different parameter condi-tions are simulated by Matlab software.The proposal for parameter selection in deep-space communication system is given based on anal-ysis on simulation results.

针对Turbo乘积码(Turbo Product Codes,TPCs)中的译码问题,该文提出一种全新的低复杂度TPC自适应Chase迭代译码算法。与已有的报道不同,在译码过程中,新算法首先统计TPC码块内每一行(列)产生的代数译码后的备选序列与接收序列的相同最小欧氏距离的个数,然后根据统计结果,按照算法步骤调整译码所需的不可靠位数值。通过Monte Carlo仿真可验证,当TPC行列编码采用相同的扩展汉明码,且编码效率为0.879时,该算法与Pyndiah采用固定不可靠位数值迭代译码算法相比,在误码率BER为410-处仅损失约0.08 dB的性能,但是译码平均复杂度降低可达到约40.4%。

This paper proposes a novel and low-complexity adaptive Chase iterative decoding algorithm for Turbo Product Codes (TPCs). Different from the previous reported results, during decoding, the new adaptive algorithm is based on the statistics of the number of the candidate sequences with the same minimum squared Euclidean distance in each row or column of TPC block firstly, and then the Least Reliable Bits (LRBs) can change according to the statistical results via the proposed steps. It can be verified by Monte Carlo simulations, when using the same extended Hamming code as TPC subcodes with coding efficiency of 0.879 and the Bit Error Rate (BER) is 410- , the coding loss of the proposed adaptive algorithm is just about 0.08 dB compared with Pyndiah’s iterative decoding algorithm using the fixed LRBs parameter in Chase decoder, but the average complexity of the proposed algorithm could be reduced about 40.4%.

脉冲位置调制(PPM)无线光通信系统易受到大气湍流影响,针对于此,将Turbo技术引入到PPM的解调和乘积码译码中。基于极大似然准则导出了PPM软检测解调方法,结合分组码的SISO译码,构建了检测解调和译码联合迭代的Turbo结构,推导了其迭代原理算法。仿真分析结果表明,误码性能可随迭代次数增加改善,其中,在大气闪烁指数0.1和误码率10-6条件下,迭代3~5次相比非迭代系统获得了0.8 dB以上的增益。

To reduce the influence of atmospheric turbulence, turbo technique was used in process of PPM demodulation and product code decoding. The soft detection demodulation method of PPM was deduced based on maximum likelihood rules, combined with SISO block decoding, the structure of Turbo demodulation and decoding was constituted, iterate algo-rithm was also given. Simulation results showed that, the performance could be improved with increasing of iterations, and under condition of 0.1 atmospheric index and BER 10-6 , with 3~5 iterations the iterate algorithm would provide more than 0.8 dB gain compared to no-iteration system.