Wireless sensor networks, consists of a group of tiny nodes, which are powered by small size batteries, which can get failed very easily. The failure of a single node may lead to the failure of an entire network in the following way. If a particular node is unaware of the failed situation of its neighbor node, it simply forwards the packet again and again as it will not receive any acknowledgment. This message passing merely wastes the precious energy of the node, which leads to its failure. This process continues which leads to the failure of the entire network. It is crucial, to identify the node which is failed, as and when it happens and to reconnect the network. In this paper, a hop count based cut detection algorithm is proposed to detect the failure of the nodes and, mobile nodes are used to reconnect the partitioned network. In HCCD at each hop, every node select the node with minimum hop count and maximum link cost, thereby the worst node can be identified as cut node and this cut can be identified before the network actually fails. Experiment results shows that hop count based cut detection outperforms the traditional distributed cut detection algorithm and reconnection using mobile nodes avoids data loss.

中文翻译：

---

识别无线传感器网络中的分区

无线传感器网络由一组微小的节点组成，这些节点由小型电池供电，很容易出现故障。单个节点的故障可能通过以下方式导致整个网络的故障。如果特定节点不知道其邻居节点的故障情况，它只会一次又一次地转发数据包，因为它不会收到任何确认。这种消息传递只会浪费节点的宝贵能量，从而导致其失败。这个过程继续下去，导致整个网络的故障。至关重要的是，在发生故障时确定发生故障的节点并重新连接网络。在本文中，提出了一种基于跳数的切割检测算法来检测节点的故障，并使用移动节点重新连接分割的网络。在每跳HCCD中，每个节点都选择跳数最小和链路成本最大的节点，从而可以将最差的节点识别为切割节点，并且可以在网络实际发生故障之前识别出该切割。实验结果表明，基于跳数的切割检测优于传统的分布式切割检测算法，并且使用移动节点重新连接避免了数据丢失。