Research On Data Provenance Methods In Sensor Network

With the continuous development of information technology,the Internet of Things has become more and more closely related to our lives.The nodes of the Internet of Things will contain a large amount of privacy and sensitive data.How to ensure that these private data are safe is a very important issue..Data provenance is a technology that traces the generation and derivative process of data,which has a good control and protection effect on confidential data in the Internet of Things.The sensor network is a computer network composed of many automatic devices distributed in space.The provenance in the sensor network needs to monitor its location information,and the network structure is relatively complicated.The traditional provenance model designed for the web can't meet the need of provenance semantics.And because the sensor network has weak computing power and small storage space,the provenance system needs to be light enough.Finally,the network structure of the sensor network is not fixed,and it is easy to change the network structure due to node failure or node counterfeiting attacks.Therefore,the provenance system needs to be able to distinguish between internal node structure changes and external attacks,and make response.The research work of this paper mainly includes the following three contents:(1)In view of the problem that the metadata definition of the traditional provenance model does not conform to the network structure of the sensor network,this paper redesigns the metadata of the provenance model.And because the description concept of the traditional provenance model cannot meet the recording requirements of the information,characteristics and attributes in the sensor network provenance,this article expands the description concept of the PROV model to improve the provenance model.On this basis,this paper uses RDF files to describe the provenance information,reduces the storage space of provenance data,and designs query and display algorithms based on SPARQL to ensure the efficiency of the query,simplify the provenance process,and make the display of provenance results clearer.(2)Aiming at the problem that the sensor network nodes have weak processing capabilities,small storage space,difficult to store a large amount of traceable data,and easy changes in the network structure,leading to provenance errors,this paper designs provenance data encoding and data based on bloom filters.Decoding algorithm.The algorithm uses Bloom filters to control the storage space of provenance information in the sensor network node and the provenance query time in the root node to the O(k)level.At the same time,in order to prevent the problem of provenance errors caused by attacks or network adjustments,this paper designs provenance verification and provenance collection algorithms to verify the network structure,and when the network structure changes,re-collect network information and issue an early warning,Which improves the scalability and security of the provenance system.(3)Aiming at the above research content,this paper designs a data provenance system and uses OMNet++ for network simulation to test the performance of the coding algorithm and compare it with other algorithms.