超点检测对于网络安全、网络管理等应用具有重要意义.由于存在着高速网络环境下海量网络流量与有限系统资源之间的矛盾,在线准确地监测网络流量是一个极大的挑战.随着多核处理器的发展,多核处理器的并行性成为算法性能提高的一种有效途径.目前,针对基于流抽样的超点检测方法存在计算负荷重、检测精度低、实时性差等问题,提出了一种并行数据流方法(parallel data streaming,简称PDS).该方法构造并行的可逆Sketch数据结构,建立紧凑的节点链接度概要,在未存储节点地址信息的情况下,通过简单地计算重构超点的地址,获得了良好的效率和精度.实验结果表明:与CSE(compact spread estimator),JM(joint data streaming and sampling method)方法相比,该方法具有较好的性能,能够满足高速网络流量监测的应用需求.
Combining the characteristics of peer-to-peer （P2P） and grid, a super-peer selection algorithm--SSABC is presented in the distributed network merging P2P and grid. The algorithm computes nodes capacities using their...
详细信息
Combining the characteristics of peer-to-peer （P2P） and grid, a super-peer selection algorithm--SSABC is presented in the distributed network merging P2P and grid. The algorithm computes nodes capacities using their resource properties provided by a grid monitoring and discovery system, such as available bandwidth, free CPU and idle memory, as well as the number of current connections and online time. when a new node joins the network and the super-peers are all saturated, it should select a new super-peer from the new node or joined nodes with the highest capacity. By theoretical analyses and simulation experiments, it is shown that super-peers selected by capacity can achieve higher query success rates and shorten the average hop count when compared with super-peers selected randomly, and they can also balance the network load when all super-peers are saturated. When the number of total nodes changes, the conclusion is still valid, which explains that the algorithm SSABC is feasible and stable.
暂无评论