Consolidation of Subtasks for Target Task in Pipelined NLP Model

Most natural language processing tasks depend on the outputs of some other tasks. Thus, they involve other tasks as subtasks. The main problem of this type of pipelined model is that the optimality of the subtasks that are trained with their own data is not guaranteed in the final target task, since the subtasks are not optimized with respect to the target task. As a solution to this problem, this paper proposes a consolidation of subtasks for a target task (CST²). In CST², all parameters of a target task and its subtasks are optimized to fulfill the objective of the target task. CST² finds such optimized parameters through a backpropagation algorithm. In experiments in which text chunking is a target task and part‐of‐speech tagging is its subtask, CST² outperforms a traditional pipelined text chunker. The experimental results prove the effectiveness of optimizing subtasks with respect to the target task.     

RTDT: A static QoS manager, RT scheduling, HW/SW partitioning CAD tool

The hardware/software partitioning/scheduling relies on two subtasks: the cost function and the real time (RT) analysis. Besides these two subtasks, the proposed generic framework, also called RT design trotter (RTDT), processes the problem of the Quality of Service (QoS) management. The aim is to add a new dimensions to solution selection, namely the guarantee of QoS from both application quality and RT issue points of view. The proposed framework defines an iteration loop of three steps that solve the sub-problems. The cost function takes into account the system on chip (SoC) area and the static and dynamic power dissipation. We show how our tool can be used to rapidly evaluate the impact of the application quality and the RT constraints choices (QoS parameters) over the final cost.     

Processor Utilization Bounds for Real-Time Systems With Precedence Constraints

This paper presents a novel approach to computing tight upper bounds on the processor utilization for general real-time systems where tasks are composed of subtasks and precedence constraints may exist among subtasks of the same task. By careful analysis of preemption effects among tasks, the problem is formulated as a set of linear programming (LP) problems. Observations are made to reduce the number of LP problem instances required to be solved, which greatly improves the computation time of the utilization bounds. Furthermore, additional constraints are allowed to be included under certain circumstances to improve the quality of the bounds.     

论提高TD-SCDMA系统覆盖能力的策略

无线覆盖是评估网络质量重要的因素之一,覆盖范围的大小直接和网络基础设施投入成正比.TD-SCDMA是我国自主研发的第三代移动通信标准,与cdma2000和WCDMA等其他制式相比有许多优点和特点,这些优点和特点使得TD-SCDMA在覆盖方面有着广阔的理论研究和实际应用前景.本文介绍了TD-SCDMA的理论覆盖半径以及影响它的覆盖能力的因素,给出了系统级仿真以及链路预算结果,提出了TDSCMDA对应的小区干扰余量算法,并从优化链路质量、降低干扰和采用覆盖技术三个方面论述如何增强TD-SCDMA系统的覆盖能力,最后提出TD-SCDMA的无线覆盖解决方案.     

Partitioning-based decoupling capacitor budgeting via sequence of linear programming

In this paper, we propose an efficient algorithm to reduce the voltage noises for on-chip power/ground (P/G) networks of VLSI. The new method is based on the sequence of linear programming (SLP) as the optimization engine, and partitioning scheme for dealing with large-sized circuits. We show that by directly optimizing the decoupling capacitor (decap) areas as the objective function and using the time-domain adjoint method, SLP can deliver much better quality in terms of decap budget than existing methods based on the merged time-domain adjoint method. The partitioning strategy further improves the scalability of the proposed algorithm and makes it efficient for larger circuits. The resulting algorithm is general enough for any P/G network. Experimental results demonstrate the advantage of the proposed method over existing state-of-the-art methods in terms of solution quality at a mild computation cost increase.     

Managing technology development for safety-critical systems

This paper presents a model that determines the optimal budget allocation strategy for the development of new technologies, for safety-critical systems, over multiple decision periods. The case of the development of a hypersonic passenger airplane is used as an illustration. The model takes into account both the probability of technology development success as a function of the allocated budget and the probability of operational performance of the final system. It assumes that the strategy is to consider (and possibly fund) several approaches to the development of each technology to maximize the probability of development success. The model, thus, decomposes the system's development process into multiple technology development modules (one for each technology needed), each involving a number of alternative projects. There is a tradeoff between development speed and operational reliability when the budget must be allocated among alternative technology projects with different probabilities of development success and operational reliability (e.g., an easily and quickly developed technology may have little robustness). The probabilities of development and operational failures are balanced by a risk analysis approach, which allows the decision maker to optimize the budget allocation among different projects in the development program, at the beginning of each budget period. The model indicates that by considering reliability in the R&D management process, the decision maker can make better decisions, optimizing the balance between development time, cost, and robustness of safety-critical systems.     

一种协同神经网络不平衡参数的差分优化方法

协同模式识别方法是协同学原理应用于模式识别领域的一种新方法,识别过程中通过调整神经网络的注意参数,能改进系统的识别性能,对协同神经网络参数在不平衡条件下进行优化,能充分利用协同神经网络的自学习能力,以提高识别效果。差分进化作为一种有效的全局近似最优解的搜索算法,具有收敛性好、速度快的特点,文中提出了一种基于差分进化的方法对协同神经网络中参数进行优化,在协同神经网络的参数空间搜索最优参数,采用了均方适应度方差的机制自适应调整搜索速度和搜索精度,克服差分进化算法参数调整困难的不足,以提高算法的寻优能力,新方法具有全局兼局部寻优能力,不易陷入局部极值,同时新方法采用约简的序参量进化参数,使优化算法能有效提高协同神经网结的效率,实际图像的分类识别结果表明,注意参数的变化会导致完全不同的识别结果,另外,本文还将新算法与平衡参数的方法、其它优化的非平衡参数的协同学习算法进行了全局优化能力的比较,采用新方法具有更快的收敛速度和更优的分类识别效果。      

MAPON: Master Antenna Passive Optical Network for the Distribution of Digital Video Broadband Services in the European ACTS-CRABS Project

The European ACTS-CRABS project addresses theproblem of delivering broadband and interactivedigital video channels over a wireless 40 GHz fixedpoint-to-multipoint cellular network connectingantennas typically placed on the top of office orapartment buildings. An important issue of thisproject is the final distribution of the signal insidethe building itself. The proposed solution is based ona passive optical network (MAPON) designed to deliverthe signal from a Master Antenna to the final usersapartment. This paper presents an overview of theMAPON solution architecture and hardware, togetherwith the results obtained on an installed testbed. Theissues of low-cost design, system bandwidthlimitations, transmitter clipping and power budget areaddressed, together with the proposed solutions. Weexperimentally measured a power budget that wouldallow the distribution to up to 32 users over lengthsof the order of 150 m.     

本科毕业设计过程管理的探索

对本科毕业设计在新形势下出现的一些问题作了分析,并在毕业设计过程管理中作了一些探索,采取了一些措施,以保证毕业设计质量.通过加强毕业设计工作力度,如：建立产、学、研教学基地,加强校企合作;建立一整套毕业设计质量监控管理制度;增加综合性、设计性实验;开展毕业设计先期基础训练等.使学生的创新能力、实践能力和创业精神得到了培养,提高了毕业设计质量,增加了学生的就业竞争力和适应社会的能力.     

Constant quality video coding using video content analysis

In the literature, several rate control techniques have been proposed to aim at the optimal quality of digitally encoded video under given bit budget, channel rate and buffer size constraints. Typically, these approaches are group-of-picture (GOP) based. For longer, heterogeneous sequences, they become unacceptably complex or struggle with model mismatches. In this paper, an off-line segment-based rate control approach is proposed for controlling the distortion variation across successive shots of a video sequence when encoding with single-layer (MPEG-4 baseline, MPEG-4 AVC) and scalable (wavelet) video codecs. Consistent quality is achieved by optimally distributing the available bits among the different segments, based on efficient rate-distortion (R-D) modelling of each segment. The individual segments are defined based on shot segmentation and activity analysis techniques. The algorithm is formulated for three different distribution models: download, progressive download and streaming. The results indicate that the proposed technique improves the quality consistency significantly, while the processing overhead compared to classical two-pass variable bit-rate (VBR) encoding is limited.     

Achieving maximum bit rate in a cognitive radio network with physical layer network coding

In this paper, the power allocation problem in a relay‐assisted cognitive radio network (CRN) is considered where the secondary users exchange information in an interweave mode on the basis of physical layer network coding. In order to enhance the capacity of CRN, using multiple‐input multiple‐output and orthogonal frequency division multiplexing has become very popular in the literature. This paper goes one step further to improve the throughput of secondary users using physical layer network coding by drawing off the transmission time. The main goal is to maximize the capacity of CRN, while keeping the total interference imposed on the primary users under a certain threshold. An optimal solution to this power allocation problem with limited relay power constraint, due to the limited budget, is derived; however, because of the high complexity of this method, an efficient suboptimal solution is also proposed.     

Reliability and Performance of Star Topology Grid Service With Precedence Constraints on Subtask Execution

The paper considers grid computing systems with star architectures in which the resource management system (RMS) divides service tasks into subtasks, and sends the subtasks to different specialized resources for execution. To provide the desired level of service reliability, the RMS can assign the same subtasks to several independent resources for parallel execution. Some subtasks cannot be executed until they have received input data, which can be the result of other subtasks. This imposes precedence constraints on the order of subtask execution. The service reliability & performance indices are introduced, and a fast numerical algorithm for their evaluation given any subtask distribution is suggested. Illustrative examples are presented.     

室外场景WLAN无线参数优化方案研究

以室外CBD商圈热点为优化对象,介绍了一种室外场景的WLAN无线参数优化解决方案。该方案采用避轻就重的思路,通过AP信道隔离、功率调整以及优化AP空口资源,达到解决室外场景下同频干扰及容量受限问题的目的。     

User interface and agent prototyping for flexible working

A collaborative work group that links workers through distributed computers is only as productive as the system itself. Software agents are emerging as a way to boost workers' productivity, performing group tasks independently. The authors use a typical business conference as an example to show how graphical models and agent prototyping empower flexible working environments     

Consistent Picture Quality Control Strategy for Dependent Video Coding

Typically, a video rate control algorithm minimizes the average distortion (denoted as MINAVE) at the cost of large temporal quality variation, especially for videos with high motion and frequent scene changes. To alleviate the negative effect on subjective video quality, another criterion that restricts a small amount of quality variation among adjacent frames is preferred for practical applications. As pointed out by , although some existing proposals can produce consistent quality videos, they often fail to fully utilize the available bits to minimize the global total distortion. In this paper, we would like to achieve the triple goal of consistent quality video, minimizing the total distortion, and meeting the bit budget strictly all at the same time on the interframe dependent coding structure. Two approaches are taken to accomplish this goal. In the first algorithm, a trellis-based framework is proposed. One of our contributions is to derive an equivalent condition between the distortion minimization problem and the budget minimization problem. Second, our trellis state (tree node) is defined in terms of distortion, which facilitates the consistent quality control. Third, by adjusting one key parameter in our algorithm, a solution in between the MINAVE and the constant quality criteria can be obtained. The second approach is to combine the Lagrange multipliers method together with the consistent quality control. The PSNR performance is degraded slightly but the computational complexity is significantly reduced. Simulation results show that both our approaches produce a much smaller PSNR variation at a slight average PSNR loss as compared to the MPEG JM rate control. When they are compared to the other consistent quality proposals, only the proposed algorithms can strictly meet the target bit budget requirement (no more, no less) and produce the largest average PSNR at a small PSNR variation.     

高速铁路TD-LTE网络覆盖方案研究

针对高速铁路TD-LTE网络覆盖展开研究,首先分析在高铁部署TD-LTE网络所面临的问题,然后通过链路预算得出在不同服务等级下的建议站间距,最后通过仿真论证覆盖方案的科学性。事实证明,该方案能有效指导高铁TD-LTE网络的规划与建设。     

任务可定点拆分的资源受限项目调度问题研究

任务可定点拆分的资源受限项目调度问题是标准资源受限项目调度问题的一个扩展.针对这一问题提出了一种离散人工蜂群求解算法.离散人工蜂群算法中采用基于任务排列的食物源编码方式,并提出了能够一种保证解的可行性和离散型的候选食物源生成方法.项目中允许定点拆分的任务首先要转换为满足先序关系的分任务,然后在通过串行调度机制生成可行调度.实算表明,该算法能够有效求解任务可定点拆分的资源受限项目调度问题,同时也看到在不影响完成质量情况下,项目通过任务拆分能够在一定程度上缩短项目工期.     

云环境下一种基于任务类型匹配的负载均衡策略研究

针对云计算环境下大量并行计算节点容易产生计算节点之间的负载不均问题,本文提出了一种基于任务类型匹配的负载均衡方案。该方案针对任务集中的多种不同长度的子任务类型情况进行判定,并对当前主流的Max-Min和Min-Min两种启发式负载均衡算法进行分析,综合其优缺点,并针对任务集的类型采用不同的算法进行任务调度。实验结果表明在该负载均衡的策略下,提出的方案具有比单一应用Max-Min或者Min-Min算法具有更好的负载均衡特性和更短的完成时间。     

Robust spatio-temporal error concealment for packet- lossy H.264 video transmission

1 Introduction With the ubiquitous use of Internet and the deployment of next generation of networks, video communications are increa- singly becoming the major services in demand. Unlike data transmission, video communication is essentially time-sensitiv…     

一种基于TOA的定位优化算法

提出了一种在非视距环境下,当基站处于不良布站情况,即基站不能如蜂窝状包围住移动台时,测电波到达时间(TOA)定位算法的一种优化方法。其主要思想为先利用三基站相交线定位方法测得一个移动台的初始值坐标,再通过初始值坐标,确定基站选择和坐标优化机制,选择两两基站最优搭配,利用最优基站的圆周线相交定位方法,经过多次测量,消除NLOS误差对测量值的影响,最终确定出移动台坐标。此方法计算量小,结构简单,经验证,此方法比原来方法有更高的测量精度。