A Study On Dynamic-clustering Routing Algorithm Based On The Event-driving Wireless Sensor Network

Because event-driving sensor network in detection emergencies is highlyreal-time, it is widely used in fire, earthquake, urban traffic, animal whereabouts,space exploration, etc. But the energy of the sensor node is limited, and the energy isneither replaceable nor renewable in harsh and hostile environment, how to extend theevent-driving wireless sensor network’s life, and save the energy consumption of thenode becomes constraints for the wireless sensor network to the practical applicationof important technical bottleneck.At present, energy consumption problems of the event-driving wireless sensornetwork are mainly caused by the following three comprehensive aspects:(1) thecluster head is unevenly distributed;(2) the rest energy of the node and the distancebetween the node and the base station are not considered when selecting the clusterhead;(3) the traditional clustering routing algorithm has unreasonable defects of thecluster’s scope and cycle when facing emergency event.Based on the existing researchliterature that proposed the energy saving protocols, on the problem of enhancing thelife cycle of wireless sensor network, the wireless sensor network has made somegratifying achievements, but hasn’t considered the influence of the above threeaspects to the node energy consumption to make the best compromise. Therefore, thisarticle studies the event-driving wireless sensor network protocols, comprehensivelyconsidering the uneven distribution of the cluster head, the unreasonable election ofcluster head, unreasonable cluster’s scope and cycle factors, based on the TEEN andEDDCR protocol, proposes a dynamic routing clustering algorithm (DCRA). The corethought of the algorithm is, nodes are dormant in initial stage after deployed. Whenone regional event happens, the node enters excitement period, calculating the scopeof noticing the nodes around and the clustering time according to the seriousness ofthe incident, building cluster within the scope of the notification, and selecting theoptimal cluster head considering the residual energy of nodes and the distance fromthe node to base station, then beginning data transmission after cluster is built. Thenthe life cycle ends, all clusters are disbanded and node enters a suspended state again.According to the situation of the events, the cluster starts to be built and the optimalcluster head is selected, in this way this process circulates.Finally, simulates the DCRA algorithm proposed in this paper by the NS2, and compares the algorithm of the classic TEEN algorithm and EDDCR algorithm withthe DCRA algorithm in the two aspects of performance in network life cycle andaverage energy consumption. The simulation results show that DCRA algorithm issuperior to the TEEN algorithm and EDDCR algorithm in performance.