Clustering Mechanism Based On Crowd Sening Research And Design

Mobile Crowd Sensing(MCS)is a new perception method of Internet of things proposed in the context of ubiquitous mobile wireless sensor technology and mobile sensing devices.With the popularity of mobile intelligent terminals,a large number of sensor nodes for long-distance transmission consume too much energy of mobile nodes and cloud servers,and increases the network transmission path.In the sensing service,a large number of lost nodes not only increase the network overhead,but also reduce the network transmission success rate.As a perception technology with human unit,how to recruit appropriate participants from the crowd as perception nodes to complete the tasks of perception and data collection efficiently and in high quality has become a problem worthy of attention.In this context,due to the limited overhead of nodes and uneven transmission distance,it is very important to conduct aggregation before data transmission.Based on these problems,this paper studies clustering mechanism,which mainly includes cluster crowd sensing low-power clustering algorithm,cluster crowd sensing network research based on lost node and task cluster head recruitment algorithm based on interactive trust.It not only involves the selection of cluster heads and the division of cluster members,but also takes the service quality assessment of cluster heads as the key condition to judge the quality of cluster crowd sensing network.Aiming at the problems in clustering mechanism,based on the existing distribution of sensing nodes,this paper combines the traditional clustering algorithm to classify the clustering of MCS.and selects the optimal cluster head according to the cluster head election standard.In order to deal with the lost nodes in the clustering network,this paper establishes routes for the lost nodes on the basis of clustering according to the specific network environment,and reduces the packet loss rate of the network.When clustering is completed,some clusters will provide low quality data or maliciously spread false information,thus reducing the quality of MCS service.This paper uses a recruitment mechanism to select clusters that are more suitable for the task and more trustworthy to provide high-quality services.The main achievements of this paper are as follows:(1)Crowd sensing low power group clustering algorithm.This section is to stabilize the user participation and reduce energy consumption,use a based on the residual energy,movement speed,direction and other fitness and node density MCLEC clustering algorithm,and improved the data transfer mode,aims to reduce the energy consumption of the user and transmission cost,increase the perception of users in the network covers all cases to participate..In this algorithm,the node density and the distance between mobile nodes are taken as the reference conditions for cluster division in swarm intelligence perception network.Among them,the selection of cluster heads will affect the overall quality of the algorithm,and the selection criteria are determined by the residual energy,speed,movement direction and other fitness of the sensing nodes.The scheme can effectively reduce the energy consumption in the network and extend the life cycle of the network.Through the comparison between experiments and various clustering algorithms,the numerical analysis proves that this algorithm can effectively reduce the energy cost of the network.(2)Clustering algorithm based on crowd sensing network of missing node group.On the basis of the low-power clustering algorithm and combined with the opportunistic characteristics of MCS network,a clubbing algorithm based on lost nodes is proposed in this section to solve the clustering situation of a large number of nodes.In the algorithm,the lost node uses the hello message to learn the status of the neighbor node,and selects the appropriate neighbor node as the intermediate node to forward the information to the destination node.When it is necessary to establish a route to transmit(send)packets,the outgoing cluster node broadcasts to the valid neighbor node.At the same time,in order to reduce the rate of change of neighbor nodes and maintain temporary routing,the selection of path measurement should consider the energy of neighbor nodes and Euclidean distance between nodes.Finally,through analysis and calculation,the path with the largest comprehensive value is the data transmission path of the temporary route.In order to verify the authenticity and effectiveness of the above algorithm,a simulation experiment is carried out.Simulation results show that the algorithm can reduce packet loss and improve the success rate of network transmission.(3)Task cluster head recruitment algorithm based on interactive trust.In the cluster of MCS network model,there is no need to communicate with each intelligent device,in order to ensure that the high quality data is transmitted,this section presents a cluster of clustering first recruitment mechanism based on interaction trust,which can effectively recruit more trusted cluster headers to perform perceptual tasks and provide data for high quality transmission.Used to evaluate the relationship between the cluster header node and the cloud server.The algorithm considers the virtual interaction between the cluster and the server for the collection of the cluster,based on these interactions,and the corresponding calculation results of the network center are used as a evaluation index to attract more reliable and reliable clusters to perform perceptual tasks.At the same time,the effectiveness of the algorithm is tested on the MCS platform,and the selection method is compared with the various recruitment methods,and the qos evaluation feedback is better.