A Posteriory Probability Decoding of Nonsystematically Encoded Block Codes

We consider the problem of symbol-by-symbol a posteriori probability (APP) decoding for information symbols of nonsystematically encoded block codes. This problem arises at soft concatenated decoding of generalized concatenated block codes. The well-known BCJR algorithm for efficient APP decoding is not able to solve the problem if it runs on the minimal code trellis of a block code. We introduce an extended trellis representation for block codes, which includes encoding information and thus makes it possible to apply the BCJR algorithm as well as trellis-based decoding in the dual code space. Complexity properties of the extended trellis are investigated.     

基于优化LT码的深空数据传输策略

针对深空通信中时延长、误码率高、信道非对称及短码长LT码冗余度高、译码成功率低等问题,提出一种基于优化LT(OLT)码的深空数据传输策略。首先通过调整度分布函数、采用新颖的数据包选择策略以及联合译码算法,提出OLT码。然后,提出基于OLT码的深空数据传输策略:发送端对数据文件进行OLT码编码并发送;在接收端,使用联合译码算法对接收到的编码包进行译码得到原始数据。仿真结果表明,与LT码相比,OLT码能够提高译码成功率并降低冗余度;与CFDP相比,该策略可以有效降低时延,提高数据传输的有效性和可靠性;在高丢包率的情况下,该策略的优势更加明显。     

两种级联空时格码方案的EXIT分析

级联空时格码方案可以有效地提高空时格码系统的性能。研究了两种级联空时格码方案:PC-STTC和ST-Turbo-TC。由于两种级联方案都使用了迭代译码方法,而外信息转移(EXIT)图是分析迭代译码性能的有利工具,主要分析比较了PC-STTC和ST-Turbo-TC的EXIT性能,比较了不同信噪比和编码多项式对于译码外信息转移特性的影响。研究结果显示,PC-STTC方案比ST-Turbo-TC方案有着更好的迭代收敛性能,同时给出了BER仿真图证明了EXIT图的分析结论。     

基于DMR标准的信道编解码研究

为了提高DMR系统基带算法的性能,分析研究了DMR标准中的3/4率网格码和二次剩余码。针对3/4率网格码,提出两种译码算法,方法 1充分利用了有限状态机的特性,方法 2将维特比译码算法中的蝶形运算进行推广改进。针对QR码,通过计算校验多项式的码重进行译码。MATLAB下的仿真结果表明,3/4率网格码的两种译码方法性能相差不大,但方法一的计算量少,易于工程实现。QR码的译码算法不仅有效地避免了计算复杂的错误位置多项式,而且提高了相同信噪比下的译码性能。     

未编码DUSTM系统的网格译码

提出了未编码DUSTM系统的网格译码。将DUSTM系统视为具有全状态转移的网格结构,采用维特比软译码算法进行信号解调。仿真结果表明该方案相比传统的基于单符号最大似然解调的DUSTM系统性能有大幅提升,且比编码DUSTM系统具有更高的频谱利用率。     

最大熵模型的树-栅格最优N解码算法

最大熵模型已被广泛应用于多种自然语言处理任务,但一些现有研究工作在解码算法上存在有待改进的地方.本文提出了一个最大熵模型的树-栅格最优N解码算法,并对算法性能进行了分析和比较.算法的另一优点在于可以在解码过程中检测并控制潜在的标注冲突.     

一种可实现零内存存取的CAVLC解码算法

在基于上下文的自适应可变长度编码(CAVLC)解码算法中,对非结构化自适应可变长度编码码表进行解码时需要反复查找码表进行码字匹配,从而导致解码速度慢和需要大量内存存取的问题。为此,提出一种可实现零内存存取的CAVLC解码算法。将CAVLC码字前缀0的个数作为一级索引,同时通过一级索引获得输入码流的可能长度。将码字后缀作为二级索引并获得码字的值,直接通过码字快速获得解码结果。对于确定的输入码字,只需通过无码表查找代码操作即可得到对应的解码输出。测试结果表明,该算法不仅可以实现零内存存取的CAVLC解码,而且其解码速度比标准算法提高了45%。     

如何构造低译码复杂度格子码

文中给出了一种大自由距低译码复杂度网格编码方案。该方案是对Hellstern提出的网格编码方案的推广。使用该网格编码方案,我们可以灵活控制译码延迟。     

结构光系统的自动标定

提出一种基于单应矩阵的摄像机自动标定算法。讨论摄像机焦距为恒定和任意变化两种情况下求解摄像机内参数的计算方法:论证空间平面诱导单应矩阵的性质,利用该性质不但能求出摄像机外参数,还可得到空间平面法向量和单应矩阵方程的比例因子。该算法在求解过程中不需要非线性迭代,可以直接获得解析解,实验表明该算法具有很好的准确性、普遍性。     

基于交比不变性的空间平面参数算法

为解决利用图像坐标计算交比可能出现计算故障或者引起计算精度下降的问题,从实际计算出发,采用N矢量表示视平面上的点和直线,建立交比的N矢量计算公式.列举一个应用实例,利用交比的不变性确定空间平面参数.该算法利用空间平面上不同的4个点的N矢量计算空间平面的3D方向和视点到平面的距离.实例表明该算法比传统算法实用、可靠.     

A kinematic analysis of the space station remote manipulator system (SSRMS)

An efficient reverse analysis of three 6-degree-of-freedom (dof) subchains of the 7-dof SSRMS is presented. The first subchain is formed by locking the seventh joint. The second subchain is formed by locking the second joint, while the third subchain is formed by locking the first joint (the grounded joint is counted as the first joint in the chain). There are a maximum of eight different arm configurations in each of the three subchains, and these were determined by employing a computer-efficient algorithm, which required the rooting of only at most quadratic polynomials. The algorithms were implemented, and the SSRMS was employed in an animated environment to perform and practice a number of useful tasks for space station servicing. The locking of the second joint has the advantage in that an operator could, at the outset, choose the orientation of the plane that contains the two longest links (the upper arm and forearm) so as to avoid collisions with obstacles. However, it has the disadvantage that when the second joint angle equals 0° or 180°, the manipulator is in a singularity configuration (a singularity analysis of the SSRMS is presented in a second article). It is interesting to note that this plane can also be oriented by specifying the first joint angle. This has the distinct advantage that the plane can be oriented arbitrarily and, in this way, the singularity is avoided.     

An efficient joint implementation of three stages for fast computation of color space conversation in image coding/decoding

This paper proposes an efficient joint implementation algorithm for computing color space conversion, quantization and discrete cosine transform (DCT) in an image coder/decoder. By combining the three stages, the proposed algorithm reduces the operation amount of computing color space conversion considerably. In the case of color sampling 4:4:4, the proposed algorithm reduces the multiplication amount by 40% and the addition amount by 42% for the conversion from RGB to YCbCr in an image coder, and reduces the multiplication amount by 60% and the addition amount by 42% for the conversion from YCbCr to RGB in an image decoder. In the cases of down-sampling 4:2:2 and 4:1:1, there are the similar results. The existing fast methods in the literatures can still be applied together with this proposed algorithm into the implementation of the international image coding standards which use the transform coding technology, such as JPEG, MPEG and H.26X, and raises the image coding/decoding speed efficiently.     

改进型LLR BP译码算法研究

短低密度校验(LDPC)码的Tanner图中通常存在环路,变量节点之间的信息不再相互独立,导致LLR BP算法译码性能的下降。针对上述问题,提出一种改进型LLR BP译码算法,推导出有环时变量节点的真实信息,利用最小均方误差准则计算出有记忆的变量节点信息的权值,通过调整变量节点信息的迭代过程降低变量节点之间信息的相关性。仿真结果表明,改进型LLR BP算法具有比LLR BP算法、归一化BP算法及偏移量BP算法更好的LDPC译码性能。     

Shortest polygonal paths in space

A classical problem of geometry is the following: given a convex polygon in the plane, find an inscribed polygon of shortest circumference. In this paper we generalize this problem to arbitrary polygonal paths in space and consider two cases: in the “open” case the wanted path of shortest length can have different start and end point, whereas in the “closed” case these two points must coincide. We show that finding such shortest paths can be reduced to finding a shortest path in a planar “channel”. The latter problem can be solved by an algorithm of linear-time complexity in the open as well in the closed case. Finally, we deal with constrained problems where the wanted path has to fulfill additional properties; in particular, if it has to pass straight through a further point, we show that the length of such a constrained polygonal path is a strictly convex function of some angle α, and we derive an algorithm for determining such constrained polygonal paths efficiently.     

Feig快速DCT算法及其处理器的体系结构设计

离散余弦变换是ＪＰＥＧ和ＭＰＥＧ中的关键技术之一。当前ＭＰＥＧ的普遍应用是在解码中使用，一个解码的例子是ＶＣＤ，ＩＤＣＴ是静止／运动补偿帧解码的关键部分。为了要得到较好的性能，ＩＤＣＴ通常被硬件逻辑实现而嵌入产品中，但该方法有一个缺陷。那就是ＩＤＣＴ的硬件逻辑实现不能完成ＭＰＥＧ所需要的其它功能，如音频解码，视频／音频的比特流可变长度解码等等。文中提出的Ｆｅｉｇ快速ＤＣＴ算法的并行顺序指令流实现Ｉ     

无线信道下基于可译集的喷泉码增量译码算法

针对无线信道中数字喷泉码BP译码算法复杂度高、增量译码效率低下的问题,提出了一种基于可译集的增量译码算法。该算法给出变量节点成功译码时似然比所需达到的合适门限值Tre的理论分析方法,将译码过程中似然比高于门限值的变量节点归入可译集,提前译出以减少计算量;另一方面,若译码失败,增加开销重新译码时可先利用已成功译出的部分变量节点简化Tanner图,只对未达到译码门限的变量节点进行迭代,进一步减少计算量,并给出了算法描述和复杂度分析。最后通过仿真表明,该算法与传统的BP译码算法性能相同,但计算量大大减少,效率显著提高。     

基于变量节点更新的梯度下降比特翻转译码算法

由于变量节点自身的可靠性度量值不会随着迭代译码过程中比特被翻转而更新,因此翻转函数值的计算并不准确,进而影响了梯度下降比特翻转算法的译码性能。在深入分析梯度下降比特翻转译码算法的基础上,提出了一种基于变量节点更新的加权梯度下降比特翻转译码算法。该算法为翻转函数引入了校验节点的可靠度外信息权重和变量节点可靠性的更新规则,从而使翻转函数值的计算更加精确。仿真结果表明,与现有的梯度下降译码算法相比,在加性高斯白噪声信道下,该算法有效提升了误码率性能。     

H.264熵解码器的研究与实现

在H.264/AVC标准中,基于上下文的自适应可变长编码(CAVLC)解码算法的复杂度较高。为此,提出一种基于熵解码算法的新型熵解码器,在对视频压缩码流实现熵解码的过程中,引入并行处理方式,并改进二叉树法。通过采用QuartusⅡ7.2版环境波形仿真和FPGA硬件实现方法进行实验,结果表明该熵解码器在硬件资源节省和解码速度方面具有较好的性能。     

Reordering space design in statistical machine translation

In Statistical Machine Translation (SMT), the constraints on word reorderings have a great impact on the set of potential translations that is explored during search. Notwithstanding computational issues, the reordering space of a SMT system needs to be designed with great care: if a larger search space is likely to yield better translations, it may also lead to more decoding errors, because of the added ambiguity and the interaction with the pruning strategy. In this paper, we study the reordering search space, using a state-of-the art translation system, where all reorderings are represented in a permutation lattice prior to decoding. This allows us to directly explore and compare different reordering schemes and oracle settings. We also study in detail a rule-based preordering system, varying the length and number of rules, the tagset used, as well as contrasting with purely combinatorial subsets of permutations. We carry out experiments on three language pairs in both directions: English-French, a close language pair; English-German and English-Czech, two much more challenging pairs. We show that even though it might be desirable to design better reordering spaces, model and search errors seem to be the most important issues. Therefore, improvements of the reordering space should come along with improvements of the associated models to be really effective.     

On viewing block codes as finite automata

Block codes are viewed from a formal language theoretic perspective. It is shown that properties of trellises for subclasses of block codes called rectangular codes follow naturally from the Myhill Nerode theorem. A technique termed subtrellis overlaying is introduced with the object of reducing decoder complexity. Necessary and sufficient conditions for trellis overlaying are derived from the representation of the block code as a group, partitioned into a subgroup and its cosets. The conditions turn out to be simple constraints on coset leaders. It is seen that overlayed trellises are tail-biting trellises for which decoding is generally more efficient than that for conventional trellises. Finally, a decoding algorithm for tail-biting trellises is described, and the results of some simulations are presented.