Research On Routing Algorithm Based On Clustering And Fuzzy Inference For Wireless Sensor Networks

Sensors are the basic equipment of wireless sensor networks.Their size is small,so that called"nodes" in the networks.However,the energy of the sensor is limited,in the most cases,whitch used in scenarios where not easy to supplement electric energy.The exhaustion of the sensors energy leads to the paralysis of the networks,which affects its practical value and economic value.So that how to improve the sensors energy efficient and the energy balance has become an important topic for WSNs.Thesis focuses on study wireless sensor network routing algorithm.In order to improve the network lifetime and the energy efficient of nodes,clustering methods and fuzzy inference models are used to study energy consumption optimization methods from the scale and distribution of network clusters and the quality of cluster head node selection.The main research contents of this paper are as follows:(1)The characteristics of LEACH,EEUC,SEP,DEEC,CHEF classic routing algorithms and the different advantages of each algorithm are elaborated descripted and compared.Also,different improved methods proposed by classic routing algorithms to save network energy consumption are summarized.In order to find the existing problem of routing algorithm and improve it.(2)The problem that the LEACH protocol has obvious differences in cluster size and exist unbalanced cluster location distribution,which results in large differences in node energy.A Distance Measurement Method Based Clustering Algorithm for Wireless Sensor Network is proposed.This algorithm apply the farthest distance method in the network preparation stage to determine K initial clustering center nodes,so that the network delineates K initial clusters.Then according to the distance between nodes,update new clustering center node.Finally,with the new cluster center node update cluster until it's not changed.Clusters formed at the beginning of the network can make the network do not need to be clustered frequently,also reduce the direct communication between nodes and the base station during operation,which consumes a lot of energy.The simulation results show that the cluster distribution of the algorithm is more balanced and can effectively extend the network life time.(3)In order to improve the quality of cluster head nodes and network energy efficient in further.A Grid Optimization Combining Fuzzy Inference Routing Protocol for Wireless Sensor Networks is proposed.This algorithm divides the network into grids and uses cluster centroid changes to alternately join nodes to improve the uneven distribution of clusters in the CHEF protocol.Through the introduction of a Fuzzy Inference System,the distance between the node and the base station,the remaining energy of nodes,and the central value of nodes are used as the input for the system.Multiple factors are used to calculate the chance value of the node to compete be cluster heads.The cluster head finally selected by this method can basically achieve the conditions of high energy,near to basement and close to the center.The simulation results show that the algorithm effectively improves the network energy utilization rate and prolongs the network life time.In summary,thesis proposeed two different algorithms to improve the energy utilization rate of nodes and extend the network life cycle in wireless sensor networks.The two different algorithms from multiple perspectives using two different algorithm models work together to extend the network life time.