多维度P2P网络任务分配算法研究  

ON MULTI-DIMENSION TASK ALLOCATION ALGORITHM IN P2P NETWORK

作　　者：戴跃发[1] 吴波[1] 张权[1] 

机构地区：[1]国防科技大学电子科学与工程学院,湖南长沙410073

出　　处：《计算机应用与软件》2010年第7期200-202,216,共4页Computer Applications and Software

摘　　要：P2P计算可以充分利用网络的闲置资源,形成强大的计算能力。针对P2P计算的任务特点和P2P网络的特性,为了更好地进行任务的分配和调度,提出了一种多维度的任务分配算法。该算法使用有效的历史信息结合当前P2P网络的特征,对现有任务进行分割,从而减少调度的消耗,达到能够很好地预测网络状态,较快地完成P2P计算任务的结果。该算法有良好的扩展性能,可用于网络计算的各个方面,适合于网络环境下多任务的分配和调度。P2P computation can make full use of the idle resources in network to form powerful computing ability.For better task allocation and scheduling,a kind of multi-dimension task allocation algorithm is presented.It is in light of the task characteristic of P2P computation and P2P network property.This algorithm uses effective historical messages and in conjunction with current P2P network property to divide existing task,so it can reduce scheduling consumption of the network,and achieve an outcome of perfectly forecasting network state and pretty fast accomplishment of P2P computation tasks.The algorithm is good in scalability and can be used in every aspect of network computation.It is much suitable for multi-task allocation and scheduling in dynamic network.

关 键 词：多维度任务分配 P2P计算 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]