The Research On Event-Driven Cluster Routing In WSN

Wireless sensor network has various data collection methods for different applications, sensor nodes in event-driven wireless sensor network start to collect data and send to sink node only when detected events, that is, just send the valuable information to sink node, so this meets the detection needs as well as energy saving.Wireless sensor network has the feature that the amount of node, limited resource on single node and energy is un-chargeable if used up. Therefore, energy is always the concernment of wireless sensor network. Clustering is an important mechanize for wireless sensor network to effectively use energy and prolong the network lifetime, so this paper has done researches on prolonging network lifetime from view of event-driven sensor network application.First current typical clustering algorithms are analyzed and differences between event-driven and time-driven sensor network are compared. Then, an event-driven based asynchronous clustering protocol is proposed. Most of existing clustering algorithms are aiming at time-driven sensor network, which suppose all nodes collect and send data at the same rate and the network energy consumption is even, so they run by round periodically. However, for event-driven sensor network, the break of events is random and instant and the network energy consumption is not even. So these synchronous clustering algorithms are not suitable for event-driven applications, Asynchronous here means network runs clustering algorithm by a probability rather than pre-establish, its core is the new events can be transmitted timely and reliably. This asynchronous clustering algorithm includes three phases: (1) clustering occasion selection algorithm, clustering probability is determined by cluster head's residual energy and data transmit rate; (2) cluster head election algorithm; (3) cluster establish algorithm. Simulation results show that this protocol ensures the real-time transmission of sensitive data, reduces the packet loss rate, and efficiently balances nodes'energy consumption and prolongs network lifetime.Finally, due to traditional communication algorithm in cluster that fix slot length won't save energy maximally, a traffic-adaptive cluster communication algorithm based TDMA for sensor network is proposed. This algorithm can adjust slot length of node adaptively according to its changeable traffic, so the energy of idle consumption, state changing and cluster meaningless detecting are reduced. Simulation results show that this algorithm can save node's energy and improve the channel utilization.