采用基本块规约的控制流检测性能优化方法

控制流检测是抵御单粒子事件的有效手段之一.目前的主流方法是采用嵌入式签名技术, 但是该技术引入的检测指令过多, 导致程序效率低下. 本文使用基本块规约的技术, 在原基本块的基础上, 选择合适的约束量重新划分基本块, 减少引入的检测指令. 与8个常见算法的性能比较表明, 该方法在软错误检测覆盖率基本不变的前提下,能有效提高目标程序效率.

针对瞬时故障的控制流容错技术研究

随着微电子器件复杂度的提高，空间辐射对于计算机程序的正确性影响正越来越明显。一般情况下，这些影响并不是永久的，而是瞬时故障。无论是太空中的信息处理系统、嵌入式实时控制系统，还是计算机集群、高性能超级计算机都可能由于错误的输出而导致灾难性的后果。 传统的可靠性系统采用抗辐射部件和冗余的硬件来达到可靠性的要求，但是其价格昂贵，性能落后于今天的商用部件（COTS）。针对COTS在容错能力上存在的不足，软件容错技术可以在不改变硬件结构的情况下，有效的提高计算机系统的可靠性。 瞬时故障在软件层面上主要表现为控制流错误和数据流错误，本文主要针对控制流错误进行容错处理。软件实现的控制流容错技术通过在编译时加入冗余的容错逻辑，在程序执行时进行控制流错误的检测和处理。 如何在保证容错能力的同时，尽量降低冗余逻辑所带来的系统开销，是控制流容错需要解决的主要问题。本文从控制流错误的基本概念，容错单元的选择，签名信息的建立，签名点和检测点的插入位置几个角度对控制流容错进行研究，主要内容有： 1．对常见的控制流容错方法进行了分析比较，对其优点和不足予以说明。 2．对控制流错误进行了分类，以此为基础，提出了基于相关前驱基本块的控制流容错方法（CFCLRB）。 3．提出了一种签名流模型，提出了基于签名流模型的控制流容错方法（CFCSF）。该方法能够对基本块间控制流错误进行检测，具有较低的时间开销、空间开销和较高的错误覆盖率。同时，该方法可以根据容错尺度的要求，灵活的插入和删除签名点与检测点，具有极强的扩展性。该方法还可以应对动态函数指针这种编译时难以确定的控制流情况。 4．基于汇编指令对上述方法予以实现，并实现了国际上常用的控制流容错方法Control Flow Checking by Software Signatures（CFCSS）和Control-flow Error Detection through Assertions（CEDA）做为对比。通过加入冗余的指令逻辑，完成了对原程序的容错功能。 5．基于PIN工具实现了对控制流错误的注入，在相同的实验环境下对CFCLRB ，CFCSF，CFCSS，CEDA进行了对比实验。实验表明, CFCLRB的时间开销为26.9%，空间开销为27.6%，相比不具容错能力的原程序，其错误覆盖率从66.50%提升到97.32%。CFCSF的时间开销为14.7%，空间开销为22.1%，相比不具容错能力的原程序，其错误覆盖率从66.50%提升到96.79%。相比CFCSS，该方法的时间开销从37.2%下降到14.7%，空间开销从31.2%下降到22.1%，错误覆盖率从95.16%提升到96.79%。相比CEDA，该方法的时间开销从26.9%下降到14.7%，空间开销从27.1%下降到22.1%，错误覆盖率仅从97.39%下降到96.79%。 最后，本文对控制流容错的未来研究方向进行了展望。

基于基本块的控制流检错方法优化研究

系统的高可靠性是研究航空航天领域的一个重要指标. 由于太空环境的特殊性, 辐射和高能粒子会造成计算机系统的出现瞬时性错误, 这种错误被称作软错误, 它对航空航天器件造成了很大的影响, 严重降低系统的可靠性. 检测和防护这种软错误是航空航天系统中的重要研究方向之一. 软错误的检测和防护包括硬件防护与检错, 软硬件混合检错以及纯软件检错等. 随着商用器件的广泛使用, 与之相配合的各种软错误软件检错方法开始得到深入的研究, 在各种软件检错方法中, 控制流检测是抵御单粒子事件的有效手段之一.目前的主流方法是采用嵌入式签名技术, 但是该技术引入的检测指令过多, 导致程序效率低下. 本文从总结控制流检测技术的共同点出发, 分析该技术导致效率低下的原因：由于基本块定义的约束导致程序中基本块过多, 进而在代码注入过程中引入过多的判断及跳转指令, 导致程序效率低下. 本文针对这种情况, 提出了一种基于源代码分析的基本块规约的方法. 该方法通过修改基本块定义的约束, 使在新的基本块定义下每个基本块能够容纳更多的指令, 减少检测指令的注入, 提高效率；并且在新的基本块定义下, 原来的控制流检错方法仍可以不加修改的直接应用于新的基本块定义上. 该方法能在不修改benchmark源代码以及控制流检测方法的基础上, 选择合适的约束量重新划分基本块, 减少引入的检测指令. 本文中使用该方法以ECCA, CFCSS和RSCFC三个控制流检错方法作为验证对象, 使用这3种控制流检错方法, 在不同的约束量作用下, 对8个常见算法的benchmark进行了软错误覆盖率测试和效率测试. 多次实验数据表明, 该方法在提高检错算法效率的同时, 能够保持软错误检错的覆盖率基本不变. 在对控制流检错算法进行优化的同时, 本文还完成了相应的控制流分析工具, 基于模拟器的错误注入和代码片段执行时间检测工具等. 有效的对优化算法进行了评估和测试.

Fourth order accurate compact scheme with group velocity control (GVC)

For solving complex flow field with multi-scale structure higher order accurate schemes are preferred. Among high order schemes the compact schemes have higher resolving efficiency. When the compact and upwind compact schemes are used to solve aerodynamic problems there are numerical oscillations near the shocks. The reason of oscillation production is because of non-uniform group velocity of wave packets in numerical solutions. For improvement of resolution of the shock a parameter function is introduced in compact scheme to control the group velocity. The newly developed method is simple. It has higher accuracy and less stencil of grid points.

Scheme Construction with Numerical Flux Residual Correction (NFRC) and Group Velocity Control (GVC)

For simulating multi-scale complex flow fields like turbulent flows, the high order accurate schemes are preferred. In this paper, a scheme construction with numerical flux residual correction (NFRC) is presented. Any order accurate difference approximation can be obtained with the NFRC. To improve the resolution of the shock, the constructed schemes are modified with group velocity control (GVC) and weighted group velocity control (WGVC). The method of scheme construction is simple, and it is used to solve practical problems.

Optimized Group Velocity Control Scheme and DNS of Decaying Compressible Turbulence of Relative High Turbulent Mach Number

To overcome the difficulty in the DNS of compressible turbulence at high turbulent Mach number, a new difference scheme called GVC8 is developed. We have succeeded in the direct numerical simulation of decaying compressible turbulence up to turbulent Mach number 0.95. The statistical quantities thus obtained at lower turbulent Mach number agree well with those from previous authors with the same initial conditions, but they are limited to simulate at lower turbulent Mach numbers due to the so-called start-up problem. The energy spectrum and coherent structure of compressible turbulent flow are analysed. The scaling law of compressible turbulence is studied. The computed results indicate that the extended self-similarity holds in decaying compressible turbulence despite the occurrence of shocklets, and compressibility has little effects on relative scaling exponents when turbulent Mach number is not very high.

Reliable Validation Based on Optical Flow Visualization for CFD Simulations

A reliable validation based on the optical flow visualization for numerical simulations of complex flowfields is addressed in this paper. Several test cases, including two-dimensional, axisymmetric and three-dimensional flowfields, were presented to demonstrate the effectiveness of the validation and gain credibility of numerical solutions of complex flowfields. In the validation, images of these flowfields were constructed from numerical results based on the principle of the optical flow visualization, and compared directly with experimental interferograms. Because both experimental and numerical results are of identical physical representation, the agreement between them can be evaluated effectively by examining flow structures as well as checking discrepancies in density. The study shows that the reliable validation can be achieved by using the direct comparison between numerical and experiment results without any loss of accuracy in either of them.

Control Of Vortex Shedding From A Square Cylinder

Small circular, square, and thin-strip cross-sectional elements are used to suppress vortex shedding from a square cylinder at Reynolds numbers in the range of 1.12 x 10(4)-1.02 x 10(5). The axes of the element and cylinder are parallel. The element's size, position, and angle of attack are varied. Measurements of the fluctuating surface pressures and wake velocities, together with smoke flow visualization, show that vortex shedding from both sides of the cylinder is suppressed and the mean drag and fluctuating lift on the cylinder is reduced if the element is installed in an effective zone downstream of the cylinder. The effective zone of the circular element is shown to be much smaller than those of the other elements. The effects of Reynolds number and blockage ratio are investigated. A phenomenon of monoside vortex shedding is observed. The role of the element's bluffness is investigated and the suppression mechanism is discussed.

VISUALIZATION RESULTS OF AXISYMMETRICAL WAKE FLOW WITH/WITHOUT ACOUSTIC EXCITATION

Visualization results demonstrate the evolution of Kelvin-Helmholtz unstable waves into vortex pairing in a separated shear layer of a blunf circular. The results with acoustic excitation are quite different from that without acoustic excitation, and the phenomenon with excitation in a separated shear layer follows the rule of Devil s staircase, which always occurs in a non-linear dynamical system of two coupling vibrators.

Attractor crisis and bursting in a fluid flow with two no-slip directions

Characteristic burtsing behavior is observed in a driven, two-dimensional viscous flow, confined to a square domain and subject to no-slip boundaries. Passing a critical parameter value, an existing chaotic attractor undergoes a crisis, after which the flow initially enters a transient bursting regime. Bursting is caused by ejections from and return to a limited subdomain of the phase space, whereas the precrisis chaotic set forms the asymptotic attractor of the flow. For increasing values of the control parameter the length of the bursting regime increases progressively. Passing another critical parameter value, a second crisis leads to the appearance of a secondary type of bursting, of very large dynamical range. Within the bursting regime the flow then switches in irregular intervals from the primary to the secondary type of bursting. Peak enstrophy levels for both types of bursting are associated to the collapse of a primary vortex into a quadrupolar state.

A high order accurate difference scheme for complex flow fields

A high order accurate finite difference method for direct numerical simulation of coherent structure in the mixing layers is presented. The reason for oscillation production in numerical solutions is analyzed, It is caused by a nonuniform group velocity of wavepackets. A method of group velocity control for the improvement of the shock resolution is presented. In numerical simulation the fifth-order accurate upwind compact difference relation is used to approximate the derivatives in the convection terms of the compressible N-S equations, a sixth-order accurate symmetric compact difference relation is used to approximate the viscous terms, and a three-stage R-K method is used to advance in time. In order to improve the shock resolution the scheme is reconstructed with the method of diffusion analogy which is used to control the group velocity of wavepackets. (C) 1997 Academic Press.

Flow similariy applied in convection-based tilt sensor

The environment temperature has inevitable effects on property of the convect ion-based tilt sensors. It not only redefines the application, but also prevents the improvement of the sensor performance. Numerical simulation of the fluid flow in the chamber of a sensor was performed and the influence of the environment temperature was studied in this paper. At zero tilt angle, the temperature distribution along the perpendicular line cross the heat source at various environment temperatures was presented. It was found that the flow varied dramatically at different environment temperatures, which would cause the output signal vary accordingly, even when the tilt angle was kept at a constant, because this device works by sensing the change of flow. At the same condition, we present the numerical results when the temperature difference across the heat source and the environment was kept at the same, in those results, it was found that the temperature difference at every point along the perpendicular line cross the heat source keep the same, this result confirms the similarity principle of nature convection. Second, A method of eliminating environment temperature infect on property of convect ion-based tilt sensor, which is based on the theory of flow similarity, is proposed. It was found that a thermal transistance can be piped on the circuit of heat source to compensate the temperature of the heat source. A compensative circuit was specially designed which can keep flow similarity by changing heat source temperature in order to eliminate the influence of environment temperature. The experiment results show that above 70% temperature drift can be eliminated by this compensative circuit.

A numerical investigation on the flow physics of koi carp (Cyprinus carpio koi) burst-and-coast swimming

Abstract—Burst-and-coast is the most common locomotion type in freely routine swimming of koi carps (Cyprinus carpio koi), which consists of a burst phase and a coast phase in each cycle and mostly leads to a straight-line trajectory. Combining with the tracking experiment, the flow physics of koi carp’s burst-andcoast swimming is investigated using a novel integrated CFD method solving the body-fluid interaction problem. The dynamical equations of a deforming body are formulated. Following that, the loose-coupled equations of the body dynamics and the fluid dynamics are numerically solved with the integrated method. The two burst modes, MT (Multiple Tail-beat) and HT (Half Tail-beat), which have been reported by the experiments, are investigated by numerical simulations in this paper. The body kinematics is predicted and the flow physics is visualized, which are in good agreement with the corresponding experiments. Furthermore, the optimization on the energy cost and several critical control mechanisms in burst-and-coast swimming of koi carps are explored, by varying the parameters in its selfpropelled swimming. In this paper, energetics is measured by the two mechanical quantities, total output power CP and Froude efficiency Fr. Results and discussion show that from the standpoint of mechanical energy, burst-and-coast swimming does not actually save energy comparing with steady swimming at the same average speed, in that frequently changing of speed leads to decrease of efficiency.

Rapid quantitative detection of Yersinia pestis by lateral-flow immunoassay and up-converting phosphor technology-based biosensor

Up-converting phosphor technology (UPT)-based lateral-flow immunoassay has been developed for quantitative detection of Yersinia pestis rapidly and specifically. In this assay, 400 nm up-converting phosphor particles were used as the reporter. A sandwich immumoassay was employed by using a polyclonal antibody against F1 antigen of Y. pestis immobilized on the nitrocellulose membrane and the same antibody conjugated to the UPT particles. The signal detection of the strips was performed by the UPT-based biosensor that could provide a 980 nm IR laser to excite the phosphor particles, then collect the visible luminescence emitted by the UPT particles and finally convert it to the voltage as a signal. V-T and V-c stand for the multiplied voltage units for the test and the control line, respectively, and the ratio V-T/V-C is directly proportional to the number of Y pestis in a sample. We observed a good linearity between the ratio and log CFU/ml of Y pestis above the detection limit, which was approximately 10(4) CFU/mI. The precision of the intra- and inter-assay was below 15% (coefficient of variation, CV). Cross-reactivity with related Gram-negative enteric bacteria was not found. The UPT-LF immunoassay system presented here takes less than 30 min to perform from the sample treatment to the data analysis. The current paper includes only preliminary data concerning the biomedical aspects of the assay, but is more concentrated on the technical details of establishing a rapid manual assay using a state-of-the-art label chemistry. (c) 2006 Elsevier B.V. All rights reserved.

Removal of dibutyl phthalate by a staged, vertical-flow constructed wetland

Two sets of small scale systems of staged, vertical-flow constructed wetlands (VFCW) were operated in a greenhouse to study the purification of dibutyl phthalate (DBP) in admeasured water. Each system consisted of two chambers in which water flowed downward in chamber I and then upward in chamber 2. The systems were intermittently fed with wastewater under a hydraulic load of 420 mm(.)d(-1). The measured influent concentrations of DBP in the experimental system were 9.84 mg(.)l(-1), while the other system was used as a control and received no DBP. Effluent concentrations of the treated system averaged 5.82 mug(.)l(-1) and were far below the Chinese DBP discharge standard of less than or equal to0.2 mg(.)l(-1). These results indicate the potential purification capacity of this new kind of constructed wetland in removing DBP from a polluted water body.