适用于空间通信的LDPC码GPU高速译码架构

鉴于目前空间通信对高速、可重配置信道译码器的需求,利用图形处理器(GPU)的并行化运算特点,提出了一种低密度奇偶校验(LDPC)码软件高速译码架构。通过优化Turbo消息传递译码(TDMP)算法节点更新运算线程块内和块间并行度、减少非规则行重造成的线程分支、降低线程对节点更新信息存储资源的访问延时以及合理量化译码器存储信息来提升译码内核函数的执行效率。并在此基础上引入异步统一计算设备构架(CUDA)流处理机制,设计优化的译码器输入输出数据传输和内核函数之间的执行调度方式以及CUDA流上的译码线程资源配置方式,最大化译码吞吐率的同时降低译码延时。在Nvidia最新的Tesla K20和GTX980平台上对国际空间数据系统咨询委员会(CCSDS)遥测标准LDPC码进行的TDMP译码实验结果表明,本架构进行10次迭代译码的吞吐率最高可达约500 Mbps,平均译码延时约为2ms左右。与现有结果相比,本架构在保持软件架构配置灵活性的同时更加有效的兼顾了译码吞吐率和延时性能。     

局部概率传播译码的非系统低密度卷积码

低密度卷积码(LDCC)是低密度码(LDPC)经由其产生矩阵的侧滑变换成为带状产生矩阵而得到的,具有LDPC和卷积码的误码性能好等一些共同特征。LDCC码的和积算法译码方法可以进一步改进为局部概率传播译码方法。采用流水线结构的局部概率传播译码方法用于LDCC码得到了接近和积算法的译码性能,而且译码延迟比LDPC大为降低,译码器更为简单,便于实时工作。     

GF（q）域上的低密度校验（LDPC）码的译码及其在深空通信中的应用

低密度校验码是一类能有效逼近香农限的好码,而高进制的LDPC码具有比二进制LDPC码更好的性能,但其译码复杂度太高不利于工程应用。本文提出了一种基于协同优化算法的低复杂度的高进制LDPC码的译码算法,并讨论了其在深空通信中的应用。     

CCSDS近地LDPC译码算法研究

针对CCSDS（空间数据系统咨询委员会）推荐的近地LDPC（低密度奇偶校验）码技术进行了研究,建立了和积译码算法、对数似然和积译码算法、最小和译码算法的数学模型,并对上述译码算法的译码复杂度和译码性能进行了仿真分析.分析结果表明,和积译码算法与对数似然和积译码算法的译码性能距离香农限1.2 dB,最小和译码算法的译码性能距离香农限1.45 dB.因此,提出基于最小和译码算法的改进算法——偏移最小和译码算法与归一化最小和译码算法,并分析了这2种译码算法的译码复杂度,同时进行了大量仿真实验.实验结果表明,当偏移因子β=0.15时,偏移最小和译码算法性能达到最优,译码性能距离香农限1.25 dB;当归一化因子α=0.741 2时,归一化译码算法的译码性能达到最优,译码性能距离香农限1.2 dB.归一化译码算法具有优异的译码性能和合理的复杂度,可以遴选作为CCSDSLDPC的译码算法用于工程实现.此外,还研究了迭代次数对译码性能的影响,结果表明,当迭代次数大于10次时,译码性能提升不再明显,故工程实现时迭代次数应设置为10次.     

一种基于矩阵分裂的QC-LDPC码Log-BP译码方法

 针对高码率的准循环低密度奇偶校验码（QC-LDPC）提出了一种新的高效的log-BP部分并行译码结构,它通过矩阵分裂,将原监督矩阵分裂成多个小的矩阵,使原本的校验节点更新运算被拆分成多次处理,有效地降低了BP迭代运算的复杂度;通过组织不同小矩阵校验节点更新运算与变量节点更新运算的先后顺序,可以使不同小矩阵的校验节点更新运算与变量节点更新运算同时进行,从而提高译码器的译码速率。该方法既适用于非规则码,也适用于规则码。实验结果表明,与现有的log-BP译码方法相比,在相同的码速率下,校验节点更新单元（CNU）与变量节点更新单元（VNU）规模总量减小1/3;在相同的硬件资源下,译码速率提高1/3,另外该方法使CNU与VNU结构趋于对称,有利于设置更少的流水线级数,获得更好的时钟性能。     

Cross-layer design of LT codes and LDPC codes for satellite multimedia broadcast/multicast services

According to large coverage of satellites, there are various channel states in a satellite broadcasting network. In order to introduce an efficient rateless transmission method to satellite multimedia broadcasting/multicast services with finite-length packets, a cross-layer packet transmission method is proposed with Luby transform （LT） codes for efficiency in the network layer and low density parity check （LDPC） codes for reliability in the physical layer jointly. The codewords generated from an LT encoder are divided into finite-length packets, which are encoded by an LDPC encoder subsequently. Based on noise and fading effects of satellite channels, the LT packets received from an LDPC decoder either have no error or are marked as erased, which can be mod- eled as a binary erasure channels （BECs）. By theoretical analysis on LT parameters and LDPC parameters, the relationships between LDPC code rates in the physical layer and LT codes word lengths in the network layer are investigated. With tradeoffs between the LT codes word lengths and the LDPC code rates, optimized cross-layer solutions are achieved with a binary search algorithm. Verified by simulations, the proposed solution for cross-layer parameters design can provide the best transmission mode according to satellite states, so as to improve throughput performance for satellite multimedia transmission.     

分组码在有记忆删除信道上的性能计算

 本文首先描述了一种有记忆删除信道模型,然后提出了分组码在该信道上的性能的迭代算法。最后以Golay(23,12,7)码为例给出了多种信道参数之下的性能曲线。     

FEC design for remote control and data transmission of aeronautic and astronautic vehicles

This paper focuses on the advanced Forward Error Correction(FEC) based on LowDensity Parity-Check(LDPC) codes for remote control and data transmission of Aeronautic and Astronautic Vehicles(AAV). It shows that the nontransparent LDPC codes with odd row weights are able to resist phase inversion with a small computational overhead. To the best of our knowledge, it is the first time to use FEC to resist phase inversion. Then, a high hardware-efficient FEC design for AAV, which is capable of correcting random errors as well as phase inversion, is proposed based on circulant decomposition with circulant size of 2s. Such a(4096, 2048) LDPC code has been adopted in a data transmission system due to its excellent error performance and hardware efficiency.     

Generalized Feedback Decoding of Convolutional Codes

The use of convolutional codes with feedback decoding is a common error-correction technique in simple communication systems. A drawback of conventional feedback decoding is the limitation to a class of self-orthogonal codes which, in general, are nonoptimum. Based on distance properties of the utilized code and test-error pattern analysis, we propose generalized feedback decoding which does not have the abovementioned limitation. This is minimum distance decoding and can be applied to any convolutional code while still maintaining its simplicity. Therefore, it has the advantage of being easily adopted in existing systems. We can use the relatively complicated Viterbi or sequential decoders in larger terminals and, with the same code, use the proposed decoders in smaller terminals; otherwise, both large and small terminals must utilize the same type of decoder. The proposed decoding scheme could also be used to simplify and speed up sequential decoding. In addition, comparisons of the proposed decoder with other feedback decoders are presented.     

基于提前停止判决的编码软信息辅助载波同步

 针对目前编码辅助载波同步算法中复杂度较高、延时大的问题,提出了引入辅助停止判决机制的编码辅助载波同步算法。在现有的编码辅助载波同步结构基础上,该算法能对环路信噪比(SNR)进行实时判定,在环路SNR满足限定条件后提前停止编码辅助载波同步迭代,而不影响译码性能。采用新的相位估计方式估计含相位噪声的载波相位,提升了该条件下的环路信噪比。仿真采用码率为1/2的低密度奇偶校验(LDPC)码作为编码方式,结果表明:在误码率为10^-5时,该算法减少了约50%的编码辅助载波同步迭代次数;在含相位噪声的信号条件下,与理想解调译码相比,性能损失不超过0.15 dB。     

Numerical investigation of transcritical liquid film cooling in a methane / oxygen rocket engine

Transcritical film cooling was investigated by numerical study in a methane cooled methane/oxygen rocket engine.The respective time-averaged Navier-Stokes equations have been solved for the compressible steady three-dimensional(3-D) flow.The flow field computations were performed using the semi-implicit method for pressure linked equation(SIMPLE) algorithm on several blocks of nonuniform collocated grid.The calculation was conducted over a pressure range of 202 650.0 Pa to 1.2×107 Pa and a temperature range of 120.0 K to 3 568.0 K.Twenty-nine different cases were simulated to calculate the impact of different factors.The results show that mass flow rate,length,diameter,number and diffused or convergence of film jet channel,injection angle and jet array arrangements have great impact on transcritical film cooling effectiveness.Furthermore,shape of the jet holes and jet and crossflow turbulence also affect the wall temperature distribution.Two rows of film arranged in different axial angles and staggered arrangement were proposed as new liquid film arrangement.Different radial angles have impact on the film cooling effectiveness in two row-jets cooled cases.The case of in-line and staggered arrangement are almost the same in the region before the second row of jets,but a staggered arrangement has a higher film cooling effectiveness from the second row of jets.     

大容量光通信系统中规则LDPC码的研究

为研究规则LDPC码在大容量光通信系统的性能,对不同列重和不同迭代次数的短规则LDPC码在大容量光信道中的误码率性能进行了数据仿真。发现确定码长的规则LDPC码,存在最优的列重,使得在相同的冗余度时,最大程度地改善大容量光通信系统的性能;此外,误码率随迭代次数的增加具有收敛的特性,LDPC码译码时应合理地选取迭代次数。     

TCM的全并行维特比译码器的设计及实现

网格编码调制(TCM)是一种将编码和调制相结合,利用信号集的冗余度来获取纠错能力的技术。而用于卷积码的维特比译码器同样可以用作基于卷积码和PSK的TCM的译码。本文提出了一种适用于TCM的全并行的维特比译码的设计方案,并在Xilinx公司的v600ebg432-8上实现。     

基于纠删码的遥测链路丢帧恢复技术

针对遥测系统在恶劣天气条件、航天器姿态调整、天线指向偏离、脉冲干扰等各种因素影响下出现的遥测信号闪断以至接收机丢帧问题,提出一种适用于闪断信道的纠删码技术,用于恢复丢失的数据帧.首先介绍LDPC（低密度奇偶校验）纠删码的编译码算法,然后构造一种以短码长的随机码为外码、IRA（不规则重复累积）LDPC码为内码的级联纠删码,最后给出纠删码的应用策略和实现纠删码所需的CCSDS（空间数据系统咨询委员会）协议扩展.通过仿真实验和分析,得到了LDPC纠删码和级联纠删码的纠删性能,得出级联纠删码与LDPC纠删码均可获得接近理想的纠删性能的结论,且级联纠删码的构造更为灵活,可适应不同的应用环境,对CCSDS协议进行适当扩展后即可支持纠删码在CCSDS遥测链路中的应用.     

软输入软输出LDPC码译码算法

提出了一种软输入软输出LDPC码译码算法,采用欧氏距离作为判决准则,通过对欧式距离信息的反对数和进行迭代处理,实现软输入软输出译码;同时,还提出了该算法的一种简化方法,以降低计算复杂度。本文提出的软输入软输出算法无需接收信号的信噪比和信道状态即可实现译码,并且在性能无明显损失的前提下降低了计算复杂度。仿真结果表明,在性能与和积译码算法性能相近的情况下,计算复杂度比和积译码算法的复杂度下降了约10%。     

纠错码与神经网络

从两个方面推广了Bruck和Blaum的工作。一方面,证明了在软判决译码的情况下,线性分组码的最大似然译码等价于一个神经网络收敛于它能量函数的全局极大状态。另一方面,对GF(p)上的线性分组码,构造了一个联想记忆神经网络,使得每一个线性分组码的码字都对应于神经网络的一个全局稳定状态。     

Decoding techniques in state estimation for dynamic systems withpast histories

States of dynamic models with a higher order memory are estimated using both a stack sequential decoding algorithm and the Viterbi decoding algorithm (VDA), without higher dimensional dynamic system representation. This results in memory reduction for state estimate implementation. It is found that state estimation with a stack sequential decoding algorithm is faster and more practical than the state estimation with the Viterbi decoding algorithm, even though the estimates obtained by the Viterbi decoding algorithm are superior     

工作参数对航空发动机轴承腔内二相流动的影响

作为航空发动机润滑系统油气二相流的重要区域,主轴承腔的工作参数对内部二相流动的影响对于发动机润滑系统设计具有重要意义。利用DPM壁面液膜模型,采用CFD方法对某型发动机轴承腔简化模型内油气二相流进行了数值计算,计算结果与现有试验数据符合良好;给出了轴承腔在不同主轴转速及不同滑油流量下油膜厚度、空气和油膜速度的分布以及出口速度变化规律。     

基于DSP平台的机载高清视频编码器设计与实现

针对无人机(UVA)航拍对高清视频的需求,设计并实现了基于数字信号处理器(DSP)平台的MPEG-4高清视频编码器,具有体积小、功耗低和可靠性高的特点.根据DSP片内资源设计了存储器分配方案,优化了编码过程中的数据流传输,降低了数据访问对编码速度的影响.针对视频编码算法中复杂度较高的部分,采用了基于预测的运动估计,减少...     

分组级短码长LT喷泉码的工程应用

针对LT喷泉码在准单向链路、高差错、链路恶劣易中断的信道环境应用时出现的运算量大、译码时延长、编码效率较低等问题,文章提出使用基于数据分块的分组级短码长喷泉码方案。其思想是将短码长LT码方案和分组级LT码方案结合起来,并且在数据子块的度序列生成过程中加入度值检测机制。实验仿真结果表明:与已提出的LT码方案相比,该方案能够有效地降低译码开销、计算代价和译码时延,提高系统的最大传输效率。