| Due to its characteristics of decentralization,immutability,traceability,openness,and transparency,blockchain technology can effectively help reducing the maintenance cost of the introduction of trusted centralized institutions in IoTs,avoiding single points of failure,and ensuring that data cannot be forged and tampered,providing effective support for the development of the IoT.Therefore,the research on combining blockchain technology with IoT collaborative sensing has become an important topic in many fields such as smart city,smart home,smart transportation,and industrial Internet of Things.However,there are serious deficiencies in terms of scalability of blockchain when designing a practical IoT collaborative sensing system,which is mainly manifested in two aspects:the scale is difficult to expand and the performance and efficiency is poor.First of all,the existing blockchain model structure is relatively simple,with low transaction throughput,and the transaction frequency is limited,which is difficult to meet the large-scale and heterogeneous data sensing requirements of the collaborative Internet of Things.Furthermore,the existing blockchain consensus algorithm is inefficient,and the sensing devices in the IoT environment have higher requirements on computation,storage,and other capabilities,which is difficult to meet the real-time and lightweight data sensing requirements of the collaborative Internet of Things.Therefore,this thesis aims to improve the performance,scalability,and security of the blockchain-based system towards IoT collaborative sensing.In view of the simple blockchain model structure,it is difficult to accommodate the access and concurrent transactions of massive IoT devices(Horizontal Scalability),as well as the high complexity of consensus algorithms,the large computational and communication overhead,hinders the acceptance of lightweight devices(Vertical Scalability),the research on key technologies of blockchain scalability has been carried out.The main research works and innovation points are presented as follows.(1)In view of the lack of scalability of the blockchain itself,the sensing efficiency and transaction processing throughput cannot cope with the massive IoT sensing tasks,this thesis proposes a main-side blockchain collaborative sensing scaling mechanism integrated with edge computing.In this mechanism,to construct a hierarchical framework with clientedge-cloud collaboration.On the basis,an efficient,secure,and scalable hybrid blockchain model is proposed,which builds multiple isomorphic side chains to improve the transaction concurrent processing capability of a single main chain,the participating nodes can independently choose the working chain according to their own anonymity and privacy requirements,thus realizing the parallel processing of collaborative sensing tasks.In addition,smart contracts are employed to implement the automatic and reliable cross-layer and crossdomain interaction between the main and side chain nodes,ensuring that the interaction process is secure and free of interference.Finally,in order to solve the problems of privacy protection and data consistency in the sensing process,this thesis proposes a collaborative sensing location privacy protection algorithm and a data consistency optimization solution.Detailed theoretical analysis and experimental results show that the proposed solution effectively improves performance and efficiency of the collaborative sensing system,reduces the sensing delay,and finally achieves a balance between system scalability,privacy,and availability.(2)In view of the large computation and storage overhead caused by the introduction of side chains to the participating nodes of the blockchain,and the multi-side-chain structure increasing the data redundancy of the main chain ledger copy in the network,this thesis integrates the traditional database sharding technology to propose a blockchain dynamic sharding mechanism based on hidden Markov model.The mechanism can effectively eliminate Byzantine failures of participating nodes and enables full sharding of the communication,computation,and storage overhead of transaction processing.Specifically,fully considering the dynamic clustering characteristics of IoT collaborative sensing,the hidden Markov model is combined to construct a dynamic transaction graph for node transactions,and an adaptive dynamic fine-grained incremental sharding strategy is proposed to improve system performance.Furthermore,a multi-stage concurrent transaction consensus protocol is proposed by introducing trusted execution environment and secret-sharing to improve the efficiency of sharded blockchain consensus.Finally,this thesis uses open source frameworks and tools to test the performance of proposed solutions.Detailed theoretical analysis and experimental results show that the proposed blockchain dynamic sharding mechanism can further improve the system performance and scalability,effectively coping with largescale sensing node access to the blockchain network,while taking into account the system security and availability.(3)A blockchain lightweight consensus mechanism based on global reputation:In view of the high complexity of the blockchain consensus algorithm,large computation and communication overhead,making it difficult to accept lightweight IoT devices,this thesis proposes a lightweight blockchain Byzantine consensus mechanism by building a global reputation model.In this mechanism,the global reputation associated with nodes is calculated and adjusted according to the transaction evaluation results of collaborative sensing data,and based on this,a secure and efficient consensus leader selection and consensus reaching solution is realized.In order to ensure the fairness and accuracy of transactions,the evaluation results will also be reached a consensus,which provides a basis on the implementation of the reward or punishment mechanism for participating nodes.Furthermore,the proposed two-stage consensus including node employment relationship and data quality assessment facilitates accurate evaluation of submitted data and securely writes transaction records to the blockchain,thereby ensuring that participating nodes can receive rewards that matching the results of submitted sensing tasks.In addition,the smart contract is redesigned in the two stages of node selection and data verification to ensure the data integrity during transmission and verification,while promoting the system automatically operation according to the established rules.Finally,this thesis designs and implements a prototype of IoT collaborative sensing system based on a lightweight consensus mechanism.Detailed theoretical analysis and experimental results show that the proposed solution has high efficiency,scalability,security,and availability at the same time,meeting the real-world sensing requirements. |
PDF Full Text Request