Do plants explore habitats before exploiting them? An explicit test using two stoloniferous herbs

We tested whether the processes of exploration and exploitation can be explicitly distinguished as plants grow and develop within a habitat using two stoloniferous clonal herbs,Hydrocotyle sibthorpioides(Umbelliferae) and Potentilla anserina(Rosaceae).Ramets were planted in four circular trays differing in diameter.One replicate from each diameter-group was sampled at intervals corresponding to plant coverage of the trays,and plant biomass allocation to leaves,stolons,and roots and internode length were quantified.For both species,at early sampling times(when the smallest trays were full),total plant biomass and ramet number were larger in the smaller trays than in the larger trays.However,this trend was reversed for plants collected at later times.For H.sibthorpioides,leaf mass ratios(leaf mass to total plant mass) were significantly greater,but stolon mass ratios(stolon mass to total plant mass) were less in the small trays than in the larger ones,particularly during the early stages of the experiment.Similarly,for P.anserina,leaf mass ratios decreased in the smaller trays but increased in the larger ones as the experiment progressed.Root mass ratios showed contrasting pattern to leaf mass ratios for both species;stolon mass ratios were significantly smaller in the smaller trays than in the larger ones,although there were no obvious patterns during the course of the experiment.In addition,for both species,internode length was shorter but the number of ramets was greater in the smaller trays at early sampling times.We conclude that plants invest greater biomass in resource-exploring organs(stolons) than in resource-exploiting organs(leaves or roots) as they initially establish in a habitat.The relatively lower plant productivity in the largest trays at early sampling times presumably reflects the cost of exploration prior to resource exploitation and utilization.     

基于时间门限值的低时延编码感知路由算法

针对大多数编码感知路由算法忽略了不同数据流到达编码节点的时间不一致问题.在已有编码感知路由算法的基础上,提出了基于等待门限值的编码感知路由算法.首先利用编码条件寻找编码节点,然后引入网络测试获得等待编码时间的门限值,最后根据实际值和门限值的关系决定是否等待.仿真结果表明:使用该方案的编码感知路由算法比仅仅考虑编码机会的路由算法在编码时延和吞吐量方面有更好的效果.     

移动Ad hoc网络中具有均衡策略的稳定性路由算法

移动Ad hoc网络是一种自组织、不依赖预设的基础设施快速展开的网络。在Ad hoc网络中,移动终端可以自由移动,从而会导致链路频繁的断裂。由于该网络动态的特性,建立稳定的路由进行数据传输显得尤为重要。提出了一种新的具有均衡策略的链路稳定性路由算法,并且将它在传统的AODV协议上实现得到了一种新的协议,即N-AODV。该算法在路由建立阶段,通过节点间的相互运动预测链路的生存时间,并且将路由稳定性与跳数之间进行均衡,选择出稳定性较高、跳数较小的路径。仿真结果表明新协议相比AODV、SSA协议具有更好的综合性能,有效地提高了网络的利用率。     

Hybrid Forwarding in Cooperative Communications

A cooperative communication scheme was developed in which the relay node performs hybrid forwarding, where the relay node will not forward data for a transmission pair unless the corresponding direct transmission fails, and the relay node will not forward data for a transmission pair unless the quality of the channel between the corresponding source and the relay is good. A performance analysis of the outage probability and the diversity-multiplexing tradeoff for the scheme shows that the forwarding mechanism efficiently reduces fading in wireless channels by achieving full diversity order of magnitude improvement. A considerable signal-to-noise ratio gain is provided by this scheme compared with existing cooperation schemes with fixed forwarding. This flexible forwarding mechanism enables the scheme to provide the optimal diversity-multiplexing tradeoff performance.     

时间-资源协调问题的一个优化模型

提出了一个具有资源约束问题的时间—资源协调问题的数学模型、在模型里，对相互冲突的项目工期与整个被消耗的资源费用是可以建立在作为价格信息——对应的Lagrange乘子基础上进行协调，而有助于控制和监控任何时刻的资源利用情况、在对应于资源约束的Lagrangian松弛表示式里，项目工期与整个资源消耗费用可以解释为一种二人对策问题．     

Insect population differentiation in response to enviromental thermal stress

Innature,somespeciesoccupyarestrictedgeo graphicrangewhilesomeothershaveawideorevencosmopolitandistribution .Somespeciesonlyemergeinmildseasonswhilesomeothershavedominatingpopulationsoverseasons .Thus,thepopulationdy namicsofspeciesvariesindifferenttemporalandspa tialniches,reflectingdifferentrequirementsforther malbudgetandhabitattraits[1— 3].Generally ,species’abilitytosurvivethermalstress ,togetherwithotherfactors,hasdetermineditsrangeandpo tentialregionaldistribution[4 ].Inthepastdecades…     

解多目标最优化问题的权衡割平面法

在现有的一些解多目标问题交互算法的基础上,提出了一种解决一般非线性多目标最优化问题的交互式算法——权衡割平面法.它每次迭代前后所得到的解都是有效解,搜索策略灵活实用,且对线性多目标最优化问题只需一次迭代就得到最优解.     

基于信任度规避风险的撮合交易模型

针对电子商务交易自动撮合的需要,提出了基于信任度的多主体撮合交易模型(TMMTM),并给出了该模型的具体算法.在TMMTM中,定量计算交易主体的信任度时,根据交易的特点将信任指标按照对交易的不同影响,分成动态和静态信任两类指标;区别于现常用的信任管理方法对总体评估结果衰减,TMMTM通过构造信任指标分值的隶属函数更细致地处理衰减问题.最后通过某区域月度撮合交易的算例说明并分析了TMMTM的使用方法及在规避交易风险、实现优质优价方面的积极作用和意义.     

装备预防性维修的综合权衡模型研究

预防性维修综合权衡就是在装备的设计阶段对整个系统以至各组成单元做出预防性维修的决策分析。建立了两种预防性维修的综合权衡模型．即可达可用度模型和预防性维修与修复性维修费用比模型,用以在装备的研制阶段对系统及其部件进行预防性维修决策做出定量分析．并利用实例对模型进行了验证。     

彩虹表密码分析算法的图形处理器优化设计与实现

设计了一种在图形处理器(GPU)上的彩虹表密钥分析算法.结合GPU单指令多线程的特点改进了Oechslin的彩虹表算法,将预处理中彩虹链的计算分别映射到GPU的单个线程,并利用预计算链提高了在线分析的效率.所使用的硬件平台GPU Tesla C1060 相对于CPU Core2 Duo 2.8 GHz,在运行速度方面,预处理提高了41.2倍(每秒110×106次DES加密),在线分析提高了3.52倍.在此系统上用1.3 GB的磁盘空间,平均2.73 s的在线分析时间以及46%的概率,成功获得了加密选择明文的40 bit DES密钥.