In time driven data-gathering, sensor nodes generate periodic data which are gathered at the base-station. Whereas in event driven data gathering, sensors remain idle until a critical event occurs and then the event information is sent quickly to the central unit. In hybrid data-gathering, the nodes switches between time and event-driven strategy. In this paper, we propose a hybrid data-gathering protocol that minimizes the critical event reporting delay, using anycasting forwarding techniques, and maximizes the network lifetime using sleep/wake scheduling. In our protocol, the sensor nodes generate periodic events and if the nodes detect a critical event, the event information is sent quickly to the central unit. We estimate the expected critical event reporting delay using stochastic approach and validate the same in a Monte-Carlo simulation. We show the effectiveness of our protocol, that minimizes the critical event reporting delay, using ns2 simulation, and compare our protocol with existing hybrid data-gathering protocols.

在时间驱动的数据收集中，传感器节点生成定期数据，这些数据在基站收集。而在事件驱动的数据收集中，传感器将保持空闲状态，直到发生严重事件为止，然后将事件信息快速发送到中央单元。在混合数据收集中，节点在时间和事件驱动策略之间切换。在本文中，我们提出了一种混合数据收集协议，该协议使用任意广播转发技术来最小化关键事件报告延迟，并使用睡眠/唤醒调度来最大化网络寿命。在我们的协议中，传感器节点生成周期性事件，如果节点检测到严重事件，则事件信息会迅速发送到中央单元。我们使用随机方法估计预期的关键事件报告延迟，并在蒙特卡洛模拟中对其进行验证。