信道重叠信号分离方法的发展与展望

单信道多个时频重叠信号分量的分离，即仅利用一个信道或阵元接收，进行多个时频重叠信号分量的分离，是一个非常具有挑战性的工作，具有非常重要的实际应用和理论研究价值．本文从单信道分量信号定义，到信号分量可分离性分析，再到分离方法，对单信道重叠信号分离这一问题的发展过程和最新研究进展，以及一些领域中的应用进行了系统的介绍，并进一步讨论了其面临的问题和挑战．     

Management of Product Architecture Modularity for Mass Customization: Modeling and Theoretical Considerations

This paper examines the management of product architecture modularity for mass customization. Product architecture modularity is concerned with system decomposition, the selection of components and how they are linked with each other without compromising system integrity. The goal of mass customization is to produce customized goods at low cost. It has enabled many companies to penetrate new markets and capture customers whose personal needs were not met by standard products and services. Mass customization is enabled through modular product architectures, from which a wide variety of products can be configured and assembled. In order to understand the implications of product architecture modularity for mass customization, the "modularization function" is explored by applying and simulating it to a hypothetical system. Based on this exercise and sources from the literature, a new model is introduced. Furthermore, theoretical and engineering management implications are also discussed     

Complex random vectors and ICA models: identifiability, uniqueness, and separability

In this paper, the conditions for identifiability, separability and uniqueness of linear complex valued independent component analysis (ICA) models are established. These results extend the well-known conditions for solving real-valued ICA problems to complex-valued models. Relevant properties of complex random vectors are described in order to extend the Darmois-Skitovich theorem for complex-valued models. This theorem is used to construct a proof of a theorem for each of the above ICA model concepts. Both circular and noncircular complex random vectors are covered. Examples clarifying the above concepts are presented.     

Design for Environment: An Environmentally Conscious Analysis Model for Modular Design

Nowadays, more and more manufacturing enterprises are working hard to improve the environmental performance of their products. Pollution may be generated through the life cycle of a product. To maximally improve a product's overall environmental performance and reduce potential environmental impacts, Design for the Environment can be used as early as the design stage for the product. For electromechanical products, modular design is a widely used design method. While traditional modularity analysis pays little attention to environmental factors, this research aims at including comprehensive environmentally conscious criteria into modularity analysis. This paper presents a semi-quantitative environmentally conscious modular analysis model in order to reduce pollution. Eight criteria are identified to capture most of potential environmental impacts of modular products, some of which are fuzzy criteria. Fuzzy Analytic Hierarchy Process is used to rank these environmental criteria, and fuzzy numbers are used to map some uncertain judgments of decision makers with crisp numbers. Structure of a product is represented as a fuzzy graph. The proposed modularity analysis includes similarity analysis and independence analysis. In the end, an example is given to illustrate the developed methodology.      

Updating a Modular Product: How to Set Time to Market and Component Quality

In the face of shorter product life cycles, designing products with modular component parts can shorten product development time and speed up the introduction of new products in the market. Utilizing stylized models, we examine the reuse/redesign, quality, speed-to-market, and marketing decisions for two consecutive generations of a multicomponent modular product. With modularity that assumes a stable product architecture, each component can be improved by incurring a design cost that is convex increasing in the level of quality. Our study generates the following insights. When development start-up (fixed) cost is negligible, it is profitable to upgrade every component part; otherwise, it is beneficial to reuse some of the existing parts without making any design improvements in order to save on development cost. In an effort to reduce product development time while maximizing profit, we found solid evidence that the productivity level in developing every component part can be a key driver of speed-to-market. Individually, a new product launch time postponement and an R&D budget increase can lead to improvements in component part quality and overall product quality, but our models show that better improvement in quality can be achieved from launch time postponement (budget increase) when product design teams have low (high) product development productivity. Finally, when the marginal cost of producing the new product is equal to that of the old product, it is optimal to remove the old product from the market and sell only the new product.     

Modularity as a Strategy for Supply Chain Coordination: The Case of U.S. Auto

Companies across industries have admired the success of Dell Computersin using modularity as part of a mass customization strategy to achieve build-to-order and a streamlined supply chain. Many companies are attempting to emulate this successful model, including the American automotive industry. This paper focuses on how the auto industry has been attempting to move to modularity, in part, motivated by a desire to build cars to order. This movement has led to major changes in supply chain practices based partly on imitation of successful keiretsu models in Japan and a move toward modules. This study finds significant impact of modularity on outsourcing, product development, and supply chain coordination based on interviews conducted with automakers and suppliers from 2000-2003. Based on our interviews, we observe that modularity has accompanied a major reorganization of the automotive supplier industry. We identify two major issues that appear to block U.S. automakers from gaining most of the advantages possible through modularity. First, most modularity activities appear to be primarily strategically cost reduction driven, leaving the potential of modularity for mass customization largely untapped. Second, the shift in industry reorganization has not been accompanied by changes in the supply chain infrastructure to encourage long-term partnerships. We contrast this to the more gradual approach used by Toyota as it incorporates modularity on a selective basis and moves to a build-to-order model     

Managing modularity of product architectures: toward an integrated theory

The management of innovation through modular product architecture strategies is gaining increasing importance for firms, both in practice and theory. Modularity refers to a new product development strategy in which interfaces shared among components in a given product architecture are specified and standardized to allow for greater substitutability of components across product families. It is argued that the degree of modularity inherent in product architectures depends on the constituent components and interfaces. This paper introduces a mathematical model, termed the modularization function, for analyzing the degree of modularity in a given product architecture. It takes into account the following variables: components; degree of coupling; and substitutability of new-to-the-firm components. The application of the modularization function is illustrated with two elevator systems from Schindler-traction and hydraulic. The comparative analysis of the elevators captures the sensitivity and dynamics of product architecture modularity created by three types of components (standard, neutral, and unique) and two types of interfaces (fundamental and optional).     

局部线性判别分析及其在掌纹识别中的应用

在线性判别分析(LDA)算法基础上强调图像类间数据的局部可分性,提出一种称为局部LDA(LLDA)的新子空间方法,并给出LLDA的图嵌入表示。针对LLDA同样存在的小样本问题,首先给出了传统适于LDA的主成分分析(PCA)预处理方法不适于LLDA算法的证明;进而提出了基于散度差判别准则(SDDC)的LLDA(SLLDA),既克服了LLDA的小样本问题,又提供了真实比较LLDA和LDA的平台。在PolyU掌纹数据库上的实验结果表明本文提出的SLLDA算法用于识别的有效性,也验证了数据局部关系对分类的重要性。     

一种针对多分量信号的复延迟型时频分布的实现方法

 复延迟型时频分布(CTD)是近年来提出的一种新型时频分布.对于单分量调频信号,CTD具有良好的时频聚集性.但是对于包含两个或两个以上分量的多分量调频信号,直接采用定义式或传统频域卷积方法实现的CTD会产生大量的互交叉项,影响了它在各领域的应用.本文对CTD的传统频域卷积实现方法进行了三方面的修正,提出了一种适合多分量调频信号的修正型CTD频域卷积实现方法.仿真结果表明,本文方法对于多分量调频信号,既保持了CTD所固有的时频聚集性高的优点又极大地抑制了不同信号分量间的互交叉项.     

模型驱动的效能评估流程柔性构建技术

大型评估任务中,评估对象、评估数据、评估任务的多样化和复杂化使得评估人员对评估软件的柔度提出了更高的要求.针对此问题文中提出了一套模型驱动的评估流程柔性构建技术,并设计实现了一套可应用此技术完成评估流程快速灵活构建的平台.文中从流程建模语言定义、组件建模语言定义、系统组件库设计、流程模型实例生成技术四个方面阐述了该项技...     

Conceptualizing Mass Media Effect

Scholars have generated evidence of a wide variety of mass media effects over almost 9 decades of research. Although each of these effects has been defined in a relatively clear manner, there has been much less conceptualizing about what constitutes a mass media effect in general. Rarely have scholars provided a formal definition of mass media effect, instead opting to provide a definition in either an ostensive or primitive manner. In this essay, a conceptualization of “mass media effect” at the most general level is synthesized from this previous definitional work. The proposed conceptualization posits 4 general kinds of mass media effects: gradual long‐term change in magnitude, reinforcement, immediate shift, and short‐term fluctuation change.     

一种构造BCM码的新方法

由Imai与Hirakawa提出的多级编码方法,可用来构造具有任意大最小平方欧氏距离的分组调制(BCM)码.一个BCM码的性能主要取决于构成它的各个成分码,恰当地选择成分码是构造一个好的BCM码的关键.文章给出了一种新方法,通过选用不同长度的成分码来构造BCM码.仿真结果表明用此方法构造出的BCM码较传统的BCM码在性能与复杂度上有明显的改善.     

Optimization of symbolic feature extraction for pattern classification

The concept of symbolic dynamics has been used in recent literature for feature extraction from time series data for pattern classification. The two primary steps of this technique are partitioning of time series to optimally generate symbol sequences and subsequently modeling of state machines from such symbol sequences. The latter step has been widely investigated and reported in the literature. However, for optimal feature extraction, the first step needs to be further explored. The paper addresses this issue and proposes a data partitioning procedure to extract low-dimensional features from time series while optimizing the class separability. The proposed procedure has been validated on two examples: (i) parameter identification in a Duffing system and (ii) classification of fatigue damage in mechanical structures, made of polycrystalline alloys. In each case, the classification performance of the proposed data partitioning method is compared with those of two other classical data partitioning methods, namely uniform partitioning (UP) and maximum entropy partitioning (MEP).     

Product spectrum matrix feature extraction and recognition of radar deception jamming

A deception jamming recognition algorithm is proposed based on product spectrum matrix (SPM). Firstly, the product spectral in the different pulse repetition interval (PRI) is calculated, and the product spectral of frequency-slow time is arranged into a two-dimensional matrix. Secondly, non-negative matrix factorisation (NMF) is used to extract the features, and further the separability of the characteristic parameters is analysed by the F-Ratio. Finally, the best features are selected to recognise the deception jamming. The experimental results show that the average recognition accuracy of the proposed deception jamming algorithm is higher than 90% when SNR is greater than 6dB.     

一种结合分离滤波器与局部二进模式的快速人眼定位方法

为了解决低质量人脸图像中的快速人眼定位问题,提出一种结合分离滤波器与局部二进模式(LBP)的快速人眼定位算法,该方法快速有效地计算多尺度局部二进模式(LBPs)值,并且利用串级的思想,结合分离滤波器与LBP,进行快速精确的人眼定位.仿真实验表明,与一般的LBP方法相比,该方法在保证人眼定位精度的同时,具有更快的计算速度,在BiolD人脸库中,在两个尺度上计算,平均每幅人脸图像的计算时间下降近40%.     

A novel product coding and recurrent alternate decoding scheme for image transmission over noisy channels

In this letter, we present a novel product channel coding and decoding scheme for image transmission over noisy channels. Two convolutional codes with at least one recursive systematic convolutional code are employed to construct the product code. Received data are decoded alternately in two directions. A constrained Viterbi algorithm is proposed to exploit the detection results of cyclic redundancy check codes so that both reduction in error patterns and fast decoding speed are achieved. Experiments with image data coded by the algorithm of set partitioning in hierarchical trees exhibit results better than those currently reported in the literature.     

On the separability of multivariable polynomials

A simple test for the separability of multivariable (MV) polynomials which occur in the study of MV networks and multi-dimensional digital filters, etc., is presented. The separability condition is given in terms of the coefficients of the MV polynomials and expressing the given polynomial, if separable, as a product of single-variable polynomials is a direct consequence of the test for separability.     

A generic framework for blind source separation in structurednonlinear models

This paper is concerned with blind source separation in nonlinear models. Special attention is paid to separability issues. Results show that separation is impossible in the general case. However, for specific nonlinear models, the problem becomes tractable. A generic set of parametric nonlinear mixtures is considered: This set has the Lie group structure (a group structure with continuous binary operation). In the parameter set, a definition of a relative gradient is given and is used to design both batch and stochastic algorithms. For the latter, it is shown how a proper use of the relative gradient leads to equivariant adaptive algorithms     

互联网流量识别中的基于标签传播的重叠社团发现算法

The increased capacity and availability of the Internet has led to a wide variety of applications. Internet traffic characterization and application i-dentification is important for network management. In this paper, based on detailed flow data collected from the public networks of Internet Service Pro-viders, we construct a flow graph to model the in-teractions among users. Considering traffic from different applications, we analyze the community structure of the flow graph in terms of community size, degree distribution of the community, commu-nity overlap, and overlap modularity. The near line-ar time community detection algorithm in complex networks, the Label Propagation Algorithm (LPA), is extended to the flow graph for application identi-fication. We propose a new initialization and label propagation and update scheme. Experimental re-sults show that the proposed algorithm has high ac-curacy and efficiency.