Research On Trustworthy Data Collection Method Of Sensor-Cloud Based On Mobile Edge Computing

As the underlying infrastructure of the Internet of Things,wireless sensor networks(WSNs)have expanded people's ability to collect external information and realized the integration of the information world and the physical world.However,the computing capacity,storage capacity,communication capacity,and energy of sensor nodes are extremely limited.How to effectively manage large-scale wireless sensor networks to give full play to their sensing performance is still a very challenging issue.In recent years,cloud computing technology has developed rapidly.The combination of sensor networks and cloud computing technologies is becoming closer and closer.Cloud computing technologies have expanded the application space of sensor networks,and also provided new ideas for solving the many problems of sensor networks.As a result,Sensor-Cloud has emerged.The Sensor-Cloud uses the powerful computing and storage capabilities of cloud computing to provide services for sensor network applications.For example,in forest danger monitoring applications,various data(such as temperature)collected in the forest are uploaded to the cloud platform for storage and analysis,and the cloud platform's powerful computing and management capabilities are used for analysis.Once dangers such as fire are detected,the sensor cloud system can respond in time.There is no doubt that the data collected by the underlying sensor network is the foundation of the sensor cloud system and all applications.However,if the collected data is questionable and untrustworthy,it will make the upper data protection and application impossible.For example,in the application of forest fire monitoring,if the sensor node transmits the wrong data when the fire occurs,it may lead to disastrous consequences.The data collected by the sensor network is not reliable.The main troubles are listed as follows: 1)Most of the sensor nodes are deployed in the harsh natural environment,coupled with their energy shortage,they are prone to failure,resulting in inaccurate or even wrong data.As pointed out in the literature,due to noise and other factors,the data collected by the sensor will deviate,and only less than 49% of the information is effective and reliable.2)Due to the weak ability of sensor nodes,they are very vulnerable to malicious attacks and physical manipulation,resulting in invalid or even misleading data.3)In the process of data transmission,there are malicious attacks.If the data is deliberately disturbed and damaged by malicious nodes,further data screening is needed.However,these problems of sensor network are that the computing power of the underlying sensor network is too weak to provide the support of computing power and storage capacity for trustworthy data collection.To solve these problems,this study intends to introduce edge computing mode(Edge Computing).The edge computing model is an extension of the cloud computing model,similar to cloud computing.Edge computing provides computing and data storage services for end network nodes,on the one hand,the main characteristics of edge computing are close to the end of the network,with certain local computing capabilities,wider geographical distribution and support for mobility.On the other hand,some of the more powerful nodes in the sensor network(such as mobile nodes)have genes that become edge nodes.Compared with cloud computing,the architecture used for edge computing is distributed and closer to the edge of the network.It concentrates data,data processing and applications are concentrated in devices at the edge of the network,rather than submitting all data to the cloud.This study intends to introduce a mobile node between the sensor network and the cloud computing,which forms an edge computing layer with fixed base stations or convergence nodes and other devices.The above problems are solved by the trustworthy data collection of mobile edge nodes.The edge node layer can serve as a bridge and bond between sensor networks and cloud computing.Depending on the local computing capabilities of the edge nodes,complex tasks that ordinary sensor nodes cannot run can be implemented here,such as trusted evaluation and data collection.The difference between this study and traditional methods is that the trustworthiness of nodes is an important consideration.The proposed mobile data collection method should not only collect as much trustworthy data as possible but also consider factors such as energy and delay.The main research work is as follows:1)Solve the problem of node trust evaluation in the Sensor-Cloud.Based on edge computing,multi-index and multi-dimensional trust evaluation are performed on nodes in the Sensor-Cloud.The evaluation of the direct trust and indirect trust of nodes is used to effectively identify the malicious nodes in the Sensor-Cloud and ensure the feasibility of the nodes and subsequent data.2)Solve the math problem of trust value utility in Sensor-Cloud system.This paper uses trust values and Euclidean distance to generate the path with the most utility values based on the trust evaluation of nodes in the network.The path scheme based on edge computing is discussed in detail,and a motion scheme for maximizing the utility value of edge nodes is designed.3)Solve trustworthy data collection for virtual forces in Sensor-Cloud systems.This paper makes full use of the node's trust value,maps the node's trust value to the virtual force received by the node,and uses the role of the virtual force to generate a credible data collection path to ensure the reliability of the data collected by the Sensor-Cloud system.In summary,this paper designs a trustworthy data collection framework for Sensor-Cloud based on edge computing mode,which uses edge nodes as mobile data collectors to realize trusted data collection under the influence of node trust evaluation,movement with maximum utility value and virtual force.The use of mobile edge nodes for trustworthy data collection solves the problem of low-level sensor computing power and internal malicious attacks facing the nodes,and effectively extends the life cycle of the network on the premise of ensuring the trust of the nodes.To verify the proposed scheme,a large number of simulation experiments are carried out in this paper.The experimental results verify the effectiveness of the scheme.