Clustering-Based Energy Efficiency Research For Wireless Sensor Networks

Wireless sensor network(WSN)is an intelligent network composed of a large number of micro-sensor nodes through self-organization.It senses and collects the information of the surrounding environment through the ubiquitous sensor nodes,and transmits the collected information in a multi-hop way.It has become a new means of information collection and processing,which is different from the past,and has become a research hotspot in academia.The research and application of WSN will have a far-reaching impact on social development and progress.Due to the limited energy carried by wireless sensor nodes,and often deployed in harsh environments and unattended places,energy constraints become a major bottleneck in wireless sensor networks.This puts forward higher requirements for the design of routing protocols.It is necessary to design energy-efficient routing protocols and make full and reasonable use of network energy.In this regard,hierarchical clustering-based routing protocols have obvious advantages and become the focus of energy-efficient routing protocols research.In this dissertation,the energy efficiency of clustering-based routing protocols for sensor networks is studied in depth.In order to improve the network energy efficiency and prolong the network life cycle,a variety of "open source and throttling" methods are proposed.The main research results include the following four aspects:According to the characteristics of routing protocols in wireless sensor networks,the characteristics and key technologies of clustering protocols are analyzed and studied.The energy consumption of cluster mechanism is analyzed in detail.Through the analysis and comparison of typical protocols,the characteristics and shortcomings of existing clustering routing protocols are pointed out,and the methods and ways to improve energy efficiency are proposed,which lays a foundation for improving and proposing new clustering routing protocols.Because wireless sensor networks are closely related to applications,for typical environmental monitoring applications,an energy efficient routing protocol UCER-FW for non-uniform clustering wireless sensor networks based on analytic hierarchy process(AHP)is proposed.Analytic Hierarchy Process(AHP)is used in cluster head election.Cluster head election is regarded as a complex multi-objective decision-making problem.The cluster head is decomposed into multi-index constraints.The hierarchical single ranking is calculated by the method of fuzzy quantization,and the hierarchical structure model is established.The eigenvectors of the judgment matrix are solved,and the weights of each node to the total objective are obtained to determine whether they are elected or not.Nodes can sense the distance from the base station,limit the competitive radius of cluster-head election,and form clusters with decreasing scale.Considering the residual energy of nodes,the problem of "hot zone" is greatly alleviated.In cluster head election,the pre-arranged cluster head rotation strategy is adopted according to the characteristics of the network,and the nodes in the cluster take turns as the cluster head according to the predetermined order,thus avoiding the energy consumption caused by frequent elections.The inter-cluster data transmission takes the form of multi-hop and considers the residual energy and communication overhead of nodes.The nodes within the one-hop distance of Sink nodes are also included in the data forwarding node set,which balances the energy consumption of nodes.The proposed method effectively balances the energy consumption of nodes,stable and reliable,and prolongs the life cycle of the network.A clustering algorithm SOL-BPSO based on binary particle swarm optimization is proposed for wireless sensor networks with environmental energy collection.Due to the limited energy of the nodes themselves,a clustering routing protocol for wireless sensor networks with energy replenishment is proposed,which uses solar energy as supplementary energy.According to the characteristics of secondary energy,the energy supply model is designed,and the whole network is divided into clusters of uniform size.In the cluster formation,a cluster formation algorithm based on binary particle swarm optimization is proposed,which transforms the network clustering into a combinatorial optimization problem.The fitness function is designed by the energy collection efficiency of the nodes,the distance from the nodes to the cluster head and the distance from the nodes to the Sink nodes.The sum of the three weighting factors is 1.Through iterative optimization,nodes with high energy efficiency are selected as cluster heads.The energy consumption of cluster nodes is minimum and balanced,and cluster heads are closer to Sink nodes.It reduces the transmission energy consumption and effectively prolongs the network life cycle.In order to solve the problem of data forwarding routing optimization in wireless sensor networks,a clustering routing algorithm based on Q-learning is proposed.In reinforcement learning,the Q-learning algorithm has the characteristic that the current state and the action's Q value generalizes all the information needed in a single value in order to determine the accumulative returns in the future when the action a is selected under the state s.This feature makes it more advantageous for routing in WSN networks.According to the characteristics of WSN,the return value is composed of residual energy of nodes,forwarding hops and transmission energy consumption between nodes.By sending feedback messages between agents to update the Q table items,the optimal next hop node is found until it is transmitted to the base station,which optimizes the path selection of data transmission and reduces the communication energy consumption.