A virtual replica node-based flash crowds alleviation method for sensor overlay networks
A virtual replica node-based flash crowds alleviation method for sensor overlay networks
复制标题
DOI：
10.1016/j.jnca.2016.09.006
复制
发表时间：
2016-11
期刊：
影响因子：
--
通讯作者：
Xun Shao;M. Jibiki;Y. Teranishi;N. Nishinaga
中科院分区：
文献类型：
--
作者：
Xun Shao;M. Jibiki;Y. Teranishi;N. Nishinaga
传感器网络的快速发展使得通过整合分散的传感资源构建大规模的传感器覆盖网络成为可能。在我们之前的工作中，我们采用了一系列可查询的 P2P（点对点）网络 Skip Graph 作为传感器覆盖网络的覆盖基底。范围可查询的 P2P（例如 Skip Graph）能够以有效且可扩展的方式检索属性在指定范围内的传感资源。范围可查询属性的副作用是，当发生闪存拥挤时，具有连续键/ID的多个节点很可能几乎同时成为热点，这在典型的基于DHT（分布式哈希表）的系统中很少发生。在这项工作中，我们提出了一种基于虚拟副本节点的方法来保护传感器覆盖网络免受闪存人群的影响。在所提出的方法中，热点节点可以请求位于覆盖网络中任何位置的物理备用节点，以使用其密钥生成虚拟节点，并强制虚拟副本节点加入覆盖网络，就像它们是热点周围的正常节点一样。该方法对于单个热点和热点区域场景均有效。通过理论分析、模拟和试验台实验，我们证明了所提出的方法是高效的、可扩展的和可行的。
The rapid development of sensor networks has made it possible to build large-scale sensor overlay networks by integrating separated sensing resources. In our previous work, we employed a range queriable P2P (peer-to-peer) network, Skip Graph, as the overlay substrate of sensor overlay networks. Range queriable P2Ps such as Skip Graph enable retrieving sensing resources whose properties are within the specified range in an effective and scalable manner. A side effect of the range queriable property is that when flash crowds occur, it is likely that multiple nodes with successive keys/IDs will become hotspots at virtually the same time, which rarely happens in typical DHT (Distributed Hash Table)-based systems. In this work, we present a virtual replica node-based approach to protect sensor overlay networks from flash crowds. In the proposed approach, a hotspot node can request physically spare nodes located anywhere in the overlay to generate virtual nodes with its key and force the virtual replica nodes to join the overlay network as if they were normal nodes around the hotspot. The proposed method is effective for both single hotspot and hotspot zone scenarios. With theoretical analysis, simulations, and testbed experiments, we demonstrate that the proposed method is efficient, scalable and feasible.