Study On Routing Algorithm Based On Data Aggregation For Wireless Sensor Networks

In recent years, wireless sensor networks have more and more widely application in theindustry and daily life. Wireless sensor network is composed of a large number of tiny sensornodes. It has a highly scalable, limited energy and limited computing power. And it is alwaysarranged in areas, where person is hard to reach. So how to use the energy efficiently and thenprolong lifetime of network become the main research questions. Considering about thewireless sensor nodes’ high distribution density and the high data correlation of the adjacentnodes, it can bring unnecessary energy and communication bandwidth consumption to theentire network without processing of all the data before transmission. Therefore, this papercombines data fusion and routing to optimize the network topology and balance networkconsumption, on the other hand, it can reduce the amount of data transmission and saveenergy consumption.This paper will discuss the wireless sensor network routing algorithm based on datafusion by the three aspects below. At the same time, the improved algorithm will be proposed.By comparing with the methods of predecessors, it will prove that the algorithms in this paperhave more advantages.1) As the existing LEACH clustering routing algorithms commonly have the problems ofuneven clustering and unbalanced overall network energy consumption, improved K-ACOclustering routing algorithm is proposed in this paper. K-medoids clustering algorithm cancluster random distributed sensor nodes, which is a good solution to the uneven clustering.Combined with improved LEACH, when replace the cluster head node and select the nexthop for data transmission between the cluster head nodes, this algorithm can consider noderesidual energy and balance network energy consumption.2) The nodes close to the base station more easily consume energy because they are in aheavier relay traffic, which is known as the hot spots problem. To solve the problem, anunequal clustering algorithm based on fuzzy for Wireless Sensor Networks is proposed in thispaper. Take the nodes’ residual energy and the distance away from base station as the input offuzzy control to get the radius of the cluster head competition. All nodes are divided into anumber of unequal clusters. The size of the clusters close to the base station is smaller thanthe clusters far from the base station. So the cluster-head close to the base station can reserveenergy for data forwarding between clusters. And this is not a static constant radius, it willconstantly adjust their competitive radius considering of the nodes’ residual energy to extendthe lifetime of the networks.3) Against the correlation between the different network nodes, the degree of data fusionis not the same,this paper develops energy-efficient data aggregation trees to reduce theenergy consumption of wireless sensor networks by using a multi-hop data aggregationapproach. First, a game theory framework is proposed to reveal the data correlation betweennodes. Subsequently, a cost function is constructed by fully considering the residual energy ofsensor nodes, exogenous disturbance and in-network data aggregation. Then, the Correlation-Aware Routing Algorithm (CARA) is employed to obtain the optimal solution,which minimize the total energy of the network.