Research On Interoperability Of Cloud Platform Towards Iot

In the era of the Internet of Everything,the collaboration of different IoT systems has become an important factor affecting the development of IoT technology.In the cloud platform of IoT,an effective data interaction mechanism can improve the interoperability between cloud platforms.Although many researchers have studied the connectivity and interoperability between cloud platforms,the interoperability problem between cloud platforms in the IoT scenario still needs to be solved.This paper aims to improve the interoperability among cloud platforms under the IoT scenario through standardization and explore the effective collaboration mechanism among multi-cloud platforms.The research contents mainly include the following aspects.Firstly,this paper analyzed the interoperability problems between the cloud platforms and ways to improve cloud platform interoperability.Based on the characteristics of the IoT scenarios,this paper designs a cloud platform that conforms to OCF specification in a standardized way.For OCF cloud platform implementation cannot resolve the data model of device resources,optimize the OCF cloud platform interoperability process,using OpenAPI on cloud services standardization description,using reusable components characteristic of the OpenAPI 3.0 data model to model of IoT equipment resources,realize the cloud platforms IoT data model of the correct analytical equipment resources.Secondly,the access control policy of the interoperation between cloud platforms is studied.In view of the access token leakage problem,the access control management mechanism of cloud platform based on alliance chain technology is designed,and the alliance chain network is implemented based on the open source project Hyperledger Fabric of the alliance chain to improve the security of user resources in the cloud platform.The consensus mechanism in distributed system was studied.Aiming at the deficiency of voting mechanism in RAFT algorithm,a priority-based RAFT voting mechanism optimization was proposed.By setting a priority for each RAFT cluster node,the election failure caused by a split vote was reduced and the leader election time was reduced.Finally,the cloud service placement algorithm is studied in the IoT scenario,the characteristics of the cloud service in the IoT are analyzed,and the cloud service placement scenarios are divided into exclusive and shared.In the case of exclusive cloud service placement,a delay-aware cloud service placement algorithm is proposed to reduce the delay loss of communication between cloud services.In the shared cloud service placement scenario,a load-based cloud service grouping merge placement algorithm is proposed to minimize the communication delay between cloud services and reduce the usage of physical machines.To solve the problem of mixed deployment of multi-type cloud services in data centers,a cloud service placement algorithm based on cosine distance was proposed,and the algorithm was optimized by greedy random adaptive search process and bubble search,which effectively reduced the amount of physical machines in data centers.