信息栅格技术及其发展

文章阐述了信息栅格的基本概念,探讨了信息栅格技术的产生动因、发展趋势及其在军事方面的应用,分析了信息栅格技术的典型———美军的全球信息栅格(GIG)。     

一种基于多Agent强化学习的多星协同任务规划算法

在分析任务特点和卫星约束的基础上给出了多星协同任务规划问题的数学模型.引入约束惩罚算子和多星联合惩罚算子对卫星Agent原始的效用值增益函数进行改进,在此基础上提出了一种多卫星Agent强化学习算法以求解多星协同任务分配策略,设计了基于黑板结构的多星交互方式以降低学习交互过程中的通信代价.通过仿真实验及分析证明该方法能...     

片上网络中一种单周期2GHz无缓冲路由器

近年来,无缓冲路由器由于不需要缓冲器而成为片上网络低开销的解决方案。为了提高无缓冲路由器的性能,提出一种单周期高性能无缓冲片上网络路由器。该路由器使用一个简单的置换网络替换串行化的交换分配器与交叉开关以实现高性能。虚通道路由器与基准无缓冲路由器相比,该路由器在TSMC65nm工艺下可以以较小的面积开销达到2GHz的时钟频率。在合成通信负载与真实应用负载下的模拟结果表明,该路由器的包平均延迟远小于虚通道路由器和其他无缓冲路由器。     

基于复杂网络的联合作战指挥体系研究

以作战指挥理论需求为牵引,从复杂网络的角度对联合作战指挥体系问题进行网络抽象,将指挥体系中的指挥机构作为网络的节点,将指挥关系作为网络的边,按照指挥体系历史发展的规律,从简单到复杂,分别构建了传统层级纵长树状型的指挥体系网络模型,在树状模型基础上,增加了横向联系但依然具有层级结构的信息共享改进型的指挥体系网络模型,以及...     

扩展信念网络模型的军事情报检索系统设计

着眼信息化战争条件下军事情报分类、处理困难的现实,在深入分析军事情报检索系统特点基础上,提出一个基于分类的自动存储和利用查询术语同义词关系扩展信念模型的检索军事情报检索原型系统,详细描述了其存储和检索设计思路.     

IED casualties mask the real problem: it's us

Casualty figures suggest that the US/Allied Counter Improvised Explosive Device (C-IED) policy and the present allocation of national assets, resources, and intellectual capital have not been very successful. A number of explanations for why this has been the case are discussed and critiqued here.     

Crime facilitation purposes of social networking sites: A review and analysis of the ‘cyberbanging’ phenomenon

There have been growing claims in media circles and law-enforcement settings that street gangs and criminal groups are turning to Internet-based social networking sites for various reasons ranging from the showcasing of their images and exploits to the suspected recruitment of members. The present study investigates whether such a trend is, in fact, in place. The presence of street gangs on these Internet sites is referred to as cyberbanging. While there is some anecdotal evidence suggesting that gangs are turning to social networking sites, there is little available research on exactly how street gangs and criminal groups use the Internet. Our main argument is that gang culture is in many ways an individualized phenomenon and this feature ties in directly with recent assessments of the Internet as a setting that is governed by a process of networked individualism. This theoretical link between the individualized gang setting and the presence of gang members on social networking sites helps us understand why recruitment is improbable even in a context where people are openly diffusing their image and exploits to a growing number of Internet users. The empirical segment of this research adds to this general outlook. Based on a keyword search of over 50 street gang names, the three main social networking sites (Twitter, Facebook, and MySpace) were monitored for street gang presence. Results illustrate that gang presence on social networking sites is linked primarily to promoting a general gang or street culture through individual displays. In regard to the visitors to such sites, there is no evidence that they are being tricked or manipulated in any way. They are, however, showing their curiosity in regard to such groups and, for those who share their comments and opinions, signs of support are evident. Thus, whereas criminal gangs are not proactively using the Internet to convert anyone into being gang members, social networking sites are creating a new venue for people who share or are sensitive to the values underlying street gang lifestyle to come together. These sites essentially create a new convergence setting for gang members to interact with a wider number of people who would probably never have been exposed to their lifestyles and exploits through physical interactions. The study's conclusion extends these findings toward further research in this area, as well as outlining the more relevant implications for law-enforcement monitoring of this growing phenomenon.     

信息推荐系统中的朋友关系预测算法设计

伴随着互联网规模的不断扩展,信息过载问题越来越突出。信息推荐系统被视为解决信息过载问题的最有效方法。然而目前的方法大多数仅考虑用户独立的反馈,而忽略用户的社会属性对推荐的重要作用,这对信息推荐系统的性能会造成巨大的影响。为此,本文提出了基于朋友关系预测的信息推荐算法,将用户的社会关系预测引入信息推荐过程中,分别基于用户的拓扑信息及历史交互信息建立用户社会关系的存在性判定及关系类型判定,并利用线性回归分析方法和逻辑回归分析方法实现了基本特征的融合。最后,通过在Epinions和Slashdot真实数据集上的实验证明,本方法能够有效提高用户社会关系预测的准确性。     

面向架次率的机场资源配置优化模型

架次出动率是评价军用飞机作战能力的一项重要指标.面向架次率这一重要指标,基于排队理论和边际分析法建立了机场资源配置优化模型.首先建立架次生成过程的闭排队网络模型,使用平均值分析方法获得给定资源配置下,任意架飞机参与飞行时的架次出动率;然后结合平均值分析算法和边际分析方法,综合考虑资源的价值和对架次率的影响,建立面向架次率的资源配置优化模型.此模型可为机场资源配置提供决策方案,并通过一个实例介绍了本模型的应用.     

基于多变元网络的Cyber作战建模方法

Cyberspace是近年出现的一个新概念,分析了Cyberspace的概念及特性,提出Cyberspace层次概念模型,在此基础上,提出基于多变元网络的Cyber作战建模方法,认为Cyber作战建模可以分为对战争系统静态结构建模和交战行为动态交互建模,并对应分别建立了Cyberspace战争系统的多变元网络模型和Cyber作战交战行为的多变元网络演化模型,最后运用所建立的模型对一个典型Cyber战进行了分析.