Research On Data Aggregation Problem In Duty-cycled Sensor Networks

In the past few years,in many applications of wireless sensor networks(WSNs),duty-cycled protocols and data aggregation are widely applied in long-term applications to reduce energy consumption.However,because of the existence of sleep latency,the aggregation delay is largely increased.In some real-time applications,e.g.fire monitoring,people are eager to obtain the aggregated data from WSNs within a short time.Hence the research on how to decrease the aggregation delay in duty-cycled WSNs is of great importance.On the other hand,batterypowered sensor nodes would be deployed in the target area for a long period of time to sense data,and the battery is not replaceable.Thus,how to reduce the energy consumption of each sensor node and prolong the network lifetime is also an important problem.The problem of minimum latency of data aggregation schedule in duty-cycled WSNs is to design a schedule method for nodes in WSNs,its purpose is to minimize the transmission delay.An efficient schedule method can not only reduce the delay,but also avoid conflicts during transmissions.The problem of prolonging the network lifetime during the process of data aggregation is to construct a aggregation tree with maximum lifetime.These works related to minimum latency of data aggregation schedule and prolonging the network lifetime in duty-cycled WSNs are described as follows:For the problem of minimum latency of data aggregation schedule,an improved method(ATC+LAS)has been put forward.This method includes two processes: data aggregation tree construction(ATC)and aggregation schedule method(LAS).ATC is based on the connected dominating set(CDS).During the construction of aggregation tree in ATC,sleep latency is considered,which is different from traditional construction method.This construction method minimizes the time for each node in aggregation tree transmitting data to the sink node.Then based on ATC,LAS uses an improved collection coverage method to handle conflicts,thus that all nodes can transmit conflict-free to the sink node.The simulation shows that the proposed scheme(ATC+LAS)needs less time to aggregate data comparing with other traditional algorithms.Under different network parameters,the proposed aggregation scheme(ATC+LAS)outperforms the state-of-the-art schemes.Finally,a theoretical analysis has been given finally,which gives an upper bound of the total aggregation delay.For the problem of prolonging the network lifetime,an improved method(LBD+LBLA)has been put forward to construct a aggregation tree.This method includes two steps: connected dominating set construction(LBD)and leaves assignment(LBLA).Both LBD and LBLA consider load balance of each node during construction to reduce nodes' average energy consumption,thus prolong the network lifetime.The simulation shows that(LBD+LBLA)can prolong the network lifetime while the total delay is within a fixed time constraint.