一种动态加权的LDPC译码方法

为了加快低密度奇偶校验（LDPC）码的译码速度，有效改善LDPC码的译码性能，针对校验节点更新过程中的对数似然比（LLR）值的大小，设计了一种LDPC码的动态加权译码方法。以IEEE 802.16e标准的奇偶校验矩阵为例，根据LLR值的变化规律，利用增长因子和抑制因子对和积译码算法和最小和译码算法进行动态加权。仿真结果显示，基于动态加权的译码方法相对于传统译码方法误码率都有明显改进，译码复杂度也有所降低。     

LT码新型度分布及其在电力线通信中的应用

Luby变换(Luby Transform,LT)码作为信道编码应用于电力线通信(Power Line Communication,PLC)，可实现电信号的可靠传输。度分布对LT码编译码性能的影响至关重要。为了得到更优的度分布，首先调整二进制指数分布(Binary Exponential Distribution,BED)中的度数比例，获得一种译码性能更优的改进的二进制指数分布(Improved BED,IBED)。然后，根据IBED在冗余度较小时译码成功率高，而冗余度增大后鲁棒孤子分布(Robust Soliton Distribution,RSD)的译码性能表现更佳的特点，通过求和归一化的方式将IBED与RSD两种度分布的优势进行有机结合，提出一种新型二进制鲁棒孤子分布(Binary RSD,BRSD)。仿真结果表明，与其他方法及传统的RSD相比，采用新度分布进行LT编码，可明显降低译码开销，并节约编译码耗时。将新型度分布应用于基于LT码的PLC系统中，能有效地抑制PLC信道中各种噪声对电信号的干扰，并提高通信效率。     

Income comparison and happiness: The role of fair income distribution

Relative income gap is one of the most popular approaches for explaining the income–happiness relationship. We argue in this article that when people compare their incomes, they care about distributional fairness more than relative income disparity. It is difficult for us to explain China's income–happiness paradox if we simply compare the income gap and do not explore the income‐generation process leading to income inequality. We therefore employ an approach based on a responsibility‐sensitive theory of justice that decomposes individual income into fair and unfair components. As a proxy for distributional unfairness, unfair income is considered the main source of unhappiness. Using data from the Chinese Household Income Project survey, we find strong support for the negative relationship between income unfairness and happiness. We also find a significantly positive relationship between the relative income gap and income unfairness, which leads us to consider the income comparison hypothesis as the explanation for the income–happiness paradox in a new light. Sensitivity analyses confirm the robustness of our results.     

基于深度学习的跳频信号识别

针对跳频信号分选存在人工提取参数特征具有复杂性的问题，提出了一种基于深度学习的识别方法。首先对跳频信号进行短时傅里叶变换，得到二维的时频矩阵；接着提取信号的轮廓特征，构造三维矩阵作等高线图，并对等高线图进行预处理；最后把预处理后的等高线图输入到卷积神经网络中进行训练、测试，进而实现分类识别。仿真结果表明，在不需要复杂的人工提取参数特征的基础上，在分选率为100〖WT《Times New Roman》〗%〖WTBZ〗时，所提方法经裁剪处理下的信噪比为-15 dB，比支持向量机和传统K-Means聚类算法都低10 dB。实测数据的算法验证表明，所提方法能够将大疆精灵4Pro、hm无人机、司马航模X8HW以及大疆悟2这四类无人机正确分类。     

利用ACE值构造QC-LDPC码

为进一步提升中短码长下准循环低密度奇偶校验(Quasi-cyclic Low-density Parity-check,QC-LDPC) 码的纠错性能，提出了一种综合短环数目和环连通性的QC-LDPC码构造方法。首先，采用Golomb规则构造QC-LDPC码，对基矩阵中的部分元素进行替换预处理，初步降低短环数目；其次，采用所提的利用近似环外信息度（Approximate Cycle Extrinsic message degree，ACE）的消环掩模算法来优化QC-LDPC码，使得掩模后的校验矩阵具有较大的ACE平均值，最终完成QC-LDPC码的构造。该构造方法简单、通用性强，在短环数目和连通性间进行了平衡。与只考虑减少短环数目、增大围长等方法相比，该方法构造的QC-LDPC码有更加优异的纠错性能。     

一种基于微波雷达回波信号的车型分类方法

车辆类型识别方法是智能交通系统的关键技术之一。利用深度学习的高维特征泛化学习能力，将改进的LeNet-5卷积神经网络用于基于交通微波雷达的大小车型分类识别。首先，以雷达触发前的N帧信号为基础，对雷达的回波信号进行分析并构建数据集；然后，分析LeNet-5卷积神经网络的特点；最后提出一种改进的LeNet-5卷积神经网络。实验结果表明，与传统的支持向量机方法相比，所提方法能够智能学习大小车的雷达时频信号特征，大小车型识别准确率达到97%以上，可为交通场景下的车型识别研究提供新的技术途径。     

基于卷积神经网络自编码器结构的空时分组传输方案

为提高现有端到端通信系统的泛化能力和可靠性，提出了一种基于卷积神经网络的空时分组码（Space-Time Block Coding，STBC）多输入多输出通信系统物理层方案。该方案将通信系统物理层表述、调制和解调过程联合起来形成端到端自编码器系统，引入多层一维卷积层，分别构建发射机和接收机，并扩展为多天线模式。为进一步提高系统可靠性，合理规划网络结构和参数，联合信号的调制和编码方案，优化了系统模型。仿真实验表明，针对瑞利相关衰落下多输入多输出（Multiple-Input Multiple-Output，MIMO）信道应用场景，训练模型可以实现传统STBC系统的误码性能，两发两收系统在发送端相关系数为0和0.9时分别优于传统系统0.5 dB和1 dB。此外，经过优化后的系统可获得采用卷积编码的性能改善效果，其两发两收不同工作方式优于传统1/2码率卷积编码STBC系统1~3 dB。     

Board diversity reforms: Do they matter for EU bank performance?

We examine the impact of governance reforms related to board diversity on the performance of European Union banks. Using a difference‐in‐difference approach, we document that reforms increase bank stock returns and their volatility within the first 3 years after their enactment. The type of reform matters, with quotas increasing return volatility. The effectiveness of reforms depends on a country's institutional environment. The impact of reforms on return volatility is found to be beneficial in countries more open to diversity, with common law system and with greater economic freedom. Finally, reforms play a bigger role in banks that have ex ante less heterogeneous boards.     

Strategies for Improving Codes of Ethics Implementation in Construction Organizations

It is necessary to reconsider the assumptions upon which the process of implementing compliance with ethical programs rests, in both theoretical and practical terms. These assumptions should hinge on organizational enablers that allow embeddedness of codes of ethics in the web of an organization's processes. This article sets out to describe an approach that will facilitate implementation of codes of ethics in construction organizations and a comprehensive literature survey approach is adopted to achieve this. The paper equally employs the application of the European Foundation for Quality Management (EFQM) model as a tool to stimulate ethical behavior in an organization, with the focus on the enabler criterion of the model. The authors discuss organizational enablers in relation to the implementation of ethical codes. The study demonstrates how ethics can be managed in an organization by proposing a framework to enhance codes of ethics embeddedness in the web of an organization. The paper indicates current research gaps and future opportunities for both academics and practitioners.     

采用航迹特征的智能机型识别

针对空战目标识别中机型自动识别比较困难的问题，提出了采用航迹特征的智能目标识别方法。利用卷积神经网络（Convolutional Neural Network，CNN）分层学习特征的能力，训练CNN算法模型自动地从航迹数据中学习有用的特征并分类。利用沿海实地采集的15个类别的飞机航迹数据，经一系列数据预处理后作为智能识别算法的训练和测试数据，在验证实验中描述了算法网络的相关配置,对比了CNN与其他分类器的识别结果。实验结果表明，CNN具有很好的识别性能。