An Event-Driven Data-Aggregation Algorithm For Wireless Sensor Networks

Data Aggregation is a crucial supporting technology of wireless sensor networks (WSN), and it mainly contains three aspects which are data aggregation at the application layer, at the network layer, and the application independent data aggregation layer. The data-aggregation algorithm at the network layer, which constructs the route from source nodes to the base station, is essential to the performance of data aggregation and the energy dissipation by nodes.One of the most important applications of WSN is monitoring emergencies, in which WSN is event-driven. However, the existing clustering data-aggregation algorithms at the network layer cluster regardless of events, so their performances are not satisfactory. Therefore, in order to save energy for wireless sensor networks applied to monitoring emergencies, an event-driven dynamic-clustering data-aggregation algorithm (EDDC) is proposed and realized. According to working mechanisms of the nervous system and dynamic-clustering data-aggregation algorithms, a novel WSN node model is established. Based on the node model, EDDC dynamically adjusts the lifetime and the scale of clusters according to events. Consequently, EDDC improves the performance of data aggregation and avoids redundant clustering. And multi-hops communication is utilized in EDDC to establish data transfer and aggregation path between source nodes and base station. This reduces energy consumption and eliminates the restriction of the scale of clusters brought by the communication range of the cluster head.The results of NS-2 simulation show that the EDDC algorithm performs better than the existing TEEN protocol in terms of average energy dissipation by nodes and quantity of nodes alive, and it considerably prolongs the lifetime of the network.