Optimal testing resource allocation during module testing considering cost, testing effort and reliability

Software reliability is one of the most important quality attributes of commercial software. During software testing, software reliability growth models (SRGMs) are commonly used to describe the phenomenon of failure occurrence and/or fault removal which consequently enhancements software reliability. Large software systems are developed by integrating a number of relatively small and independent modules, which are tested independently during module testing phase. The amount of testing resource available is limited which is desired to be consumed judiciously so as to optimize the testing process. In this paper we formulate a resource allocation problem of minimizing the cost of software testing under available amount of testing resource, given a reliability constraint. We use a flexible SRGM considering testing effort which, depending upon the values of parameters, can describe either exponential or S-shaped failure pattern of software modules. A systematic and sequential Algorithm is proposed to solve the optimization problem formulated. Numerical examples are given to illustrate the formulation and solution procedures. Sensitivity analysis is performed to examine the behavior of some parameters of SRGM with most significant influence.     

基于TMR-EHW的高可靠性电机控制系统

为提高恶劣环境中控制系统的可靠性,将三模块冗余（TMR）容错与演化硬件（EHW）自修复相结合,实现基于TMR—EHW结构的现场可编程门阵列电机控制系统。该系统利用TMR快速发现和定位故障,屏蔽出错模块并保持容错运行,利用EHW进行自修复,使出错模块恢复正常工作,系统可靠性得到提高。     

RSA算法在DSP芯片上的加速实现

RSA是一种应用广泛的公钥算法,其原理简单,安全性好,但加密速度较慢。用硬件实现数据保护在安全性和速度上都能得到更好的性能。该文介绍了一种有效实现硬件数据加密的方案--基于TMS320C54x芯片的PC安全卡,并通过分析TMS320C54x芯片的特点,提出了一种在PC安全卡上加速实现RSA算法的方法,取得了良好的加速效果。     

基于小波网络和多模块网络的数字识别

本文研究一种新的数字识别方法,这种方法用小波神经网络抽取特征、用多模块结构神经网络作模式分类器。小波分解的函数近似能力和人工神经网络的学习能力结合起来形成的小波神经网络,有着良好的特征描述性能,可用作特征抽取工具。多模块结构的神经网络将一个k类的模式分类问题转换为k个互相独立的2类分类问题。这种结构将一个复杂的分类问题化解为多个简单的分类问题,各个模块互相并联,各自负责一种模式的识别。用这种修改过的多模块结构网络的BP训练方法,可加速训练和提高训练精度,并且各模块可互相独立地进行训练。用美国NIST数字样本进行训练及测试,结果良好。这种方法可用于更广泛的平面图形识别。     

Networked reinforcement learning

Recently, many models of reinforcement learning with hierarchical or modular structures have been proposed. They decompose a task into simpler subtasks and solve them by using multiple agents. However, these models impose certain restrictions on the topological relations of agents and so on. By relaxing these restrictions, we propose networked reinforcement learning, where each agent in a network acts autonomously by regarding the other agents as a part of its environment. Although convergence to an optimal policy is no longer assured, by means of numerical simulations, we show that our model functions appropriately, at least in certain simple situations. This work was presented in part at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

高性能容错处理核心的研究

本文提出了以三模冗余表决为基础构造容错处理核心的方案。三个ＣＰＵ模块采取松散同步的形式,只有在发生总线操作时才进行同步表决。在处理核心中采用了两级故障检测机制,以三模表决器作为第一级来保证处理核心有足够高的故障覆盖率;以ＣＰＵＡ模块的自检作为第二级来弥补表决器发现故障延迟的不足,缩短故障的潜伏时间。在表决器的设计中采取了表决器旁路及先编码、后表决的方法。先用编码的方式提取需要表决的数据的特征,然后     

μReTi:小尺寸实时嵌入式系统的TCP/IP精简化研究

通过分析TCP/IP协议的不同简化实现,提出TCP/IP协议的精简策略与设计准则。采用基于模块化的跨层设计思想,实现一套TCP/IP协议简化版本—μReTi。μReTi采用静态的内存预分配与管理机制优化系统数据结构,采用宽松的分层机制以及"后门"技术减少数据复制并重用共享内存。通过在不同架构感知节点上移植μReTi,评测分析μReTi的性能,验证其低资源消耗的特点。     

Considering Prework on Industrial Projects

Prefabrication, preassembly, modularization, and off-site fabrication, collectively termed as prework have become more viable with recent advances in design and information technologies. These construction methods offer a substantial opportunity to improve project performance when circumstances merit. Successful implementation of these methods on a particular project requires systematic analysis and early decision making based on specific factors of the project. This paper identifies those factors influencing decisions on the use of prework, and current industry practices for evaluating the applicability of prework on industrial projects. It then presents a decision framework to assist industry practitioners with evaluating the applicability of prework on their project, and describes a computerized tool to aid project teams in the decision-making process. The developed framework and tool are useful and effective in the decision-making process, and easy to use, as validated by practitioners in the industry.     

A modular algorithm for resource allocation

This paper concerns resource allocation in distributed message passing systems, i.e., the scheduling of accesses to exclusive system resources shared among concurrent processes. An efficient modular resource allocation algorithm is presented that uses any arbitrary resource allocation algorithm as a subroutine. It improves the performance of the subroutine by letting each process wait only for its currently conflicting processes, and therefore, allows more concurrency. For appropriate choices of the subroutine, we obtain resource allocation algorithms with the minimum worst case response times. Simulation studies were conducted which also indicate that on average, the obtained algorithms perform faster and require a smaller number of messages than other previously known algorithms, especially when resource contention among processes is high and the average time that a process remains in the critical region is large. Received: May 1997 / Accepted: May 1998