Research On Low Energy Consumption And Secure Routing Based On K-Means++ And Hybrid Trust Model For WSN

Science and technology are progressing with the development of the times,and various technologies have come into being.The rapid progress of semiconductor technology,wireless communication technology,super-large-scale integration and Micro Electrical-Mechanical,it has paved the way for development of Wireless Sensor Network(WSN).Due to the complexity of the environment and the limitations of the node itself,the energy consumption of the node is prominent.Therefore,the issue of energy consumption has become the focus of research at home and abroad.With the continuous of the WSN,it has been widely used in health,military,national security and other fields,because of the privacy of these information,the WSN is also required to have high security.This paper takes energy consumption and routing security as the research goal,and studies related theories and improvement algorithm put forward by experts and scholars in recent years,the LEACH algorithm in the hierarchical routing algorithm with energy advantages in WSN has been improved,and propose a new routing algorithm which uniform clustering and multi-hop between clusters based on K-Means++ algorithm,then on the basis of the algorithm,a hybrid trust model is added to build trust mechanism between nodes,which improved the safety and stability of WSN with low energy consumption.The LEACH algorithm perform not well in uneven clustering and randomness of cluster head's election,in this paper,in the clustering stage,the base station uses K-Means++ algorithm to divide the cluster evenly according to the position of sensor nodes and implement the load balancing of network.Then,then the improved cluster head election formula is used to choose cluster head according to the distance from cluster center,the distance from BS and the residual energy of node,it can balance the energy consumption in a cluster.Finally,the Dijkstra algorithm is selected to establish the shortest path between clusters,it saves the energy of node and prolongs the life cycle of the network when cluster head node transfer data to the BS.The routing protocols of WSN may have many security problems,the paper proposes a hybrid trust model based on previous.Integrating the advantages of centralized and distributed trust model,the paper establishes a hybrid trust model.The cluster head nodes manage their own members in their cluster,it calculate the trust value of the cluster members according to the communication process between the cluster members,and add to the penalty factor and correction factor to ensure the accuracy of trust value.When the cluster head node's residual energy is less than the threshold,it revises the competitiveness of the cluster members by their trust value to ensure that a reliable candidate cluster head is recommended to base station.The base station calculates the cluster head's trust value through the sending or forwarding table which send out by cluster heads,and isolates the malicious cluster head timely to promises the secure multi-hop routing among cluster heads.In this paper,experiment simulation is used to verify the performance of the proposed algorithm,the results show that the proposed algorithm has improved significantly in the aspect of uniform clustering,energy load balancing and network life cycle than the original algorithm.The established hybrid trust model can detect malicious nodes quickly and isolate them timely,which improves the fault tolerance and robustness of the network.