移动IPv6网络基于身份的层次化接入认证机制

设计了一种基于身份的层次化签名方案,并在该方案基础上提出了一种适用于移动IPv6网络环境的层次化接入认证方法.该方法使用分级NAI(Network Access Identifier)作为公钥,简化了无线移动环境中的密钥管理;利用层次化思想对接入认证和移动注册进行层次化管理,减少了切换认证处理流程;基于签名机制实现了用户与接入网络的双向认证.作者用设计的切换延时分析模型,对该方法和几种传统方法进行了比较,证明当移动节点远离家乡域及在一定范围内频繁微移动时,该方法比传统方法的效率更高.通过安全性分析证明了该方法在一定程度上实现了私钥的保密性、签名的不可伪造性等功能.最后还讨论了该方法的一种可扩展变形,用于实现多级层次化移动IPv6框架下的接入认证.     

链下通道路由算法综述

链下通道网络是实现区块链系统性能提升的有效手段,其中路由算法是保证链下通道网络长期高效、稳定运行的关键.提出了链下通道网络层次化架构及路由算法基础模型;从单路路由与多路路由两方面,系统梳理了代表性路由算法及其发展过程;从有效性、并发性、可扩展性、通道平衡、路由选择中心化、成本效益、隐私保护、吞吐量、处理延迟、成功率、搜索效率11个方面设计了链下通道路由算法评价体系,对代表性路由算法进行了对比分析,并探讨了未来研究趋势.     

Bibliometric characteristics of highly cited papers from Taiwan, 2000–2009

The present study analyzes bibliometric characteristics of Taiwan??s highly cited papers published from 2000 to 2009. During this period, Taiwan ranked within the top 30 countries by number of highly cited papers, defined in Thomson Reuters?? Essential Science Indicators (ESI) as those that rank in the top 1?% by citations for their category and year of publication. Taiwan made notable progress in world-class research in the two consecutive 5-year periods 2000?C2004 and 2005?C2009. For the group of highly cited papers from Taiwan, USA, China, Germany, and Japan were the top collaborating countries over the decade. In recent years, Taiwan has increasingly collaborated with European countries whose output of highly cited papers is relatively high and increasing, rather than with its neighboring countries in Asia. Overall, Taiwan produced highly cited papers in all the 22 ESI subject categories during the 10-year period. Taiwan??s output of highly cited papers was greatest in the categories of Engineering, Clinical Medicine, and Physics, while those in Agricultural Sciences and Mathematics exceeded the expected output level in relative terms. More detailed analyses would be useful for a holistic understanding of Taiwan??s research landscape and their progress in world-class research, combining both bibliometric and non-bibliometric data, such as researcher mobility, research grants, and output from internationally-collaborated research programs.     

Synthesis and characterization of the molecularly imprinted mesoporous silica based on the self-assembly technique for selective recognition of lomefloxacin in aqueous solution

Based on double-imprinting (surfactant and lomefloxacin templates) concept and self-assembly technique, a new molecularly imprinted mesoporous silica (MIP-MS) was prepared to understand and characterize the mechanism of fast binding and selective recognition of lomefloxacin (LOM) by the adsorption isotherms model, the adsorption kinetic models, and the competitive adsorption. The MIP-MS was prepared by using LOM and surfactant micelles as the templates formed through self-hydrolyzed, self-condensed, and co-condensation of tetraethoxysilane and methyltriethoxysilane in alkaline media, and with the covalently anchored organic groups in the mesoporous silica matrix. Both the molecularly imprinted polymer (MIP) and non-imprinted mesoporous polymer (NIP) were prepared for comparison. The surface area, pore size and pore volume of the MIP-MS are 270 m² g^?1, 4.42 nm and 0.28 cm³ g^?1, respectively. The kinetics studies showed that the pseudo-second-order model was consistent with the kinetic data of the MIP-MS with the fast binding templates (the 95 % uptake of LOM within 10 min). The equilibrium data, at various temperatures, were described successfully by the Langmuir and Freundlich isotherm models. The Freundlich model was found to fit the experiment data well. The thermodynamics parameters (positive values of ΔS, negative values of ΔH and ΔG) indicated that the binding system for the MIP-MS was entropy-gained, exothermic, and the spontaneous adsorptive forces were stronger on MIP-MS than on NIP and MIP. The MIP-MS showed fast binding, higher affinity, and selectivity for the template of LOM compared with the NIP and MIP.     

Fuel-cell sharing for a distributed hybrid power system

This paper investigates the benefits of sharing a proton exchange membrane fuel cell (PEMFC) in a distributed hybrid power system. The PEMFC is usually used as backup power in stationary hybrid power systems; however, in that scenario, it might be working only 2% of the time while incurring 20% of the system expenses. Therefore, this paper examines the potential of sharing a PEMFC among multiple power systems. We develop a distributed hybrid power system that comprises several immovable power stations and a fuel-cell vehicle (FCV). Each power station is equipped with solar panels and batteries, while the FCV contains a PEMFC module and can move among the stations to provide sustainable power as needed. We propose power management strategies and show that the total system costs can be significantly reduced by 10.83% and 17.89% when sharing one FCV between three and twelve power stations, respectively. We also design experiments to demonstrate the feasibility of the proposed distributed hybrid power system. In the future, the developed model can be extended to provide further cost reductions by optimizing distributed hybrid power systems with multiple FCVs.     

某金银矿石浮选—氰化联合流程选矿试验

某金银矿石含金4.80 g/t,含银565 g/t,金矿物主要为自然金,呈独立矿物形式存在,部分以显微或次显微状赋存于褐铁矿及石英中,裸露金占总金的72.29%,铁矿物和硅酸盐矿物包裹金分别占总金的22.08%和5.63%;银矿物主要以辉银矿形式存在,嵌布粒度较粗,硫化银占总银的91.33%,自然银和氯化银含量较低,分别仅占总银的6.55%和2.12%。为实现该金银矿的高效开发利用,进行了选矿试验。结果表明：矿石在磨矿细度为-0.074 mm占74.5%的情况下,采用1粗4精2扫、精选1尾矿精扫选后再返回的浮选流程处理,可获得金品位为77.19 g/t、银品位为11 302 g/t,金、银回收率分别为75.58%和94.02%的精矿;金、银品位分别为1.23 g/t和35.45 g/t的浮选尾矿在再磨细度为-0.043 mm占86%的情况下氰化浸出,浸渣金、银品位分别为0.10 g/t和17.88 g/t,金、银对浮选尾矿的浸出率分别为91.87%和49.56%,全流程金、银总回收率分别达98.01%和96.98%。     

Research evaluation of research-oriented universities in Taiwan from 1993 to 2003

Publications have been regarded as the most significant output indicating the research performance of universities. This paper uses ISI Essential Science Indicators (ESI) database to investigate the academic performance of research-oriented universities in Taiwan, adopting the bibliometric method from both quantitative and qualitative perspectives. The data cover the time span for 11 years from 1993 to 2003. The performance indicators applied in this study includes the number of papers, the number of citations, the average citations per paper, the number of highly cited papers, the number of hot papers, and the number of top papers. The research performance and the strength of those universities are revealed in this study, and it is found that National Taiwan University leads among these universities though each university still shows strengths in various specific fields.     

Vibration behavior of light metals: Al–Zn alloy and Mg–Al–Zn alloy

Both Al–Zn alloy and Mg–Al–Zn alloy are considered as high damping and light materials. This study investigated the effects of the structure and the composition on the vibration fracture characteristics of the Al–xZn alloy and the Mg–xAl–Zn alloy under resonance. For Al–xZn (x = 7, 11, 49, and 83 wt.%) alloys, under both constant force and initial-deflection conditions, the 7Zn specimen showed the greatest vibration life. For Mg–xAl–Zn (x = 3, 6, and 9 wt.%) alloys, experimental results indicate that the AZ31-F as-extruded samples showed a greater vibration life, while high Al fully annealed samples (AZ91-O) had greater vibration fracture resistance under constant force conditions.     

基于多模态表征的移动应用GUI模糊测试框架

GUI模糊测试在提升移动应用可靠性和兼容性方面发挥着关键作用.然而,现有的GUI模糊测试方法大多效率较低,主要原因是这些工作过于粗粒度,仅基于单一模态的特征来整体理解GUI页面,应用状态的过度抽象使得许多细节信息被忽略,导致对GUI状态及小部件的理解不足.为了解决上述问题,本文提出了一种基于多模态表征的移动应用GUI模糊测试框架GUIFuzzer.该框架通过考虑多模态特征,如视觉特征、布局上下特征以及细粒度的元属性特征,来联合推断GUI小部件的语义,然后训练一个多层次奖励驱动的深度强化学习模型来优化GUI事件选择策略,提高模糊测试的效率.我们在大量的真实应用上对所提框架进行评估.实验结果表明,与现有的竞争性基线相比,GUIFuzzer显著地提升了模糊测试的覆盖率.我们还对特定目标的定制化搜索即敏感API触发进行了案例研究,进一步验证了GUIFuzzer框架的实用性.     

一种基于两级缓存的协同缓存机制

信息中心网络（information-centric networking,简称ICN）将网络通信模式从当前的以地址为中心转变为以信息为中心.泛在化缓存是ICN重要特性之一,它通过赋予网络任意节点缓存的能力来缓和服务器的压力,降低用户访问延迟.然而,由于缺少内容热度的分布感知,现有ICN缓存策略仍存在缓存利用率较低、缓存位置缺乏合理规划等问题.为了解决这些问题,提出一种基于两级缓存的协同缓存机制（a cache coordination scheme based on two-level cache,简称CSTC）.将每个节点的缓存空间分为热度感知和协作分配两部分,为不同热度的内容提供不同的缓存策略.同时,结合提出的热度筛选机制和路由策略,降低了缓存冗余,实现了缓存位置优化.最后,基于真实网络拓扑的仿真实验表明,CSTC在次热门内容缓存数量上提升了2倍,缓存命中率提升了将近50%,且平均往返跳数在多数情况下优于现有On-path缓存方式.