A blockchain-based Internet of Things method and system based on lightweight consensus algorithm

Abstract

The present invention discloses a blockchain-based Internet of Things method and system based on a lightweight consensus algorithm; the method is as follows: 1. Register a subgroup management node and a subgroup ordinary node to the Internet of Things system. 2. Set three roles: ordinary members, validators, and block producers. The subgroup management node serves as the validator; the ordinary node of the subgroup serves as the ordinary member. The block producer is elected by the validator; the rights and interests of the validator are updated and the block producer is re-elected in each working cycle. 3. The operations of all nodes in the Internet of Things layer correspond to the generation of Ethereum transactions; each block producer takes turns to package several Ethereum transactions into blocks, and enter the blockchain after consensus. The present invention realizes trusted management of the Internet of Things based on blockchain technology, reduces the single point of failure and network performance bottleneck of centralized trust, and improves the scalability of the system. The blockchain provides reliable distributed data authentication and evidence storage for Internet of Things devices.

Description

Technical Field

[0001] The present invention belongs to the technical field of the Internet of Things and blockchain, and specifically relates to a blockchain-based Internet of Things system and method based on a lightweight consensus algorithm. Background Art

[0002] The Internet of Things (IoT) has been a hot word in the field of information technology in recent years. With the continuous development and maturity of technologies such as wireless communications and sensors, mankind has entered the era of the Internet of Everything. There are a large number of nodes in the IoT system deployed at the edge of the network and on the user side that can respond to end-user requests in a timely manner. Compared with cloud computing, it is more helpful to shorten the data and service processing delay. However, the IoT has the characteristics of loose topology, frequent node joining and exiting, and dynamic distribution and autonomy of entities, which makes its security and privacy management more complex and difficult than traditional environments. At the same time, the large amount of terminal data processed by the IoT system often involves health, finance, location or other privacy data, and has strong service security and reliability requirements.

[0003] Trust is an effective alternative means of security in a distributed network environment. However, the IoT system, which is completely distributed and autonomous and has limited node resources, cannot directly use the traditional centralized and complex trust management strategy. Blockchain is a new decentralized infrastructure and distributed computing paradigm that has emerged with the application of new digital currencies such as Bitcoin. Blockchain's distributed, decentralized, secure and robust access control, as well as transparent and traceable high-quality features, provide a new path to achieve interactive reliability assurance in the IoT environment.

[0004] However, the cost of blockchain to achieve trusted interaction and distributed transaction evidence is very high, especially the core technology consensus algorithm of blockchain (proof of work PoW, proof of stake PoS, etc.) consumes a lot of energy, which is not suitable for the IoT environment with limited node computing power and real-time processing. Using blockchain technology to build a trusted IoT system also requires solving two major problems:

[0005] 1. How to manage the identity credibility of distributed devices in the Internet of Things system.

[0006] 2. Reduce the energy consumption of the trusted IoT system consensus and reduce the transaction delay caused by blockchain consensus.

[0007] At present, building a decentralized trust management model with blockchain and developing low-energy IoT trusted service interactions are still in their infancy and have great research value. Summary of the invention

[0008] The present invention provides a trusted Internet of Things architecture based on blockchain technology, proposes a QV-DPoS low-energy consumption consensus algorithm based on the reinforcement learning Q-Learning algorithm, and develops an Internet of Things identity authentication system with a blockchain architecture based on Ethereum.

[0009] In the first aspect, the present invention provides a blockchain architecture Internet of Things method based on a lightweight consensus algorithm, wherein the Internet of Things system adopted by the present invention includes a cloud server layer and an Internet of Things layer. The Internet of Things layer is divided into multiple Internet of Things subgroups. Each Internet of Things subgroup consists of a management node and multiple subgroup ordinary nodes.

[0010] The IoT approach includes the following steps:

[0011] Step 1: Register the subgroup management node and subgroup common nodes to the IoT system.

[0012] Step 2: Set three roles: ordinary members, validators, and block producers. The validator is the subgroup management node; the ordinary member is the subgroup ordinary node. The block producer is elected by the validator. In the initial state, each validator has the same rights and interests.

[0013] Each work cycle updates the rights of validators and re-elects block producers. The process is as follows:

[0014] 2-1. Update the rights and interests of all validators.

[0015] 2-1-1. Take the equity of the updated validator in the previous work cycle as the initial equity. Take the subgroup management nodes recommended as block producers by the updated validator in the previous work cycle as the feature set S.

[0016] 2-1-2. Determine whether the feature set S is an empty set; if so, it means that the validator's rights update is completed and go to step 2-2. If not, go to step 2-1-3.

[0017] 2-1-3. Take a validator in the feature set S as the identified node; determine whether the identified node becomes a block producer; if so, proceed to step 2-1-4; if not, delete the identified node from the feature set S and return to step 2-1-2.

[0018] 2-1-4. Determine whether the identified node has become an honest node; if so, proceed to step 2-1-5; if not, reduce the stake of the updated validator by z1; and delete the identified node from the feature set S, returning to step 2-1-2.

[0019] 2-1-5. Determine whether the identified node is the updated validator itself. If so, increase the stake of the updated validator by z2, delete the identified node from the feature set S, and return to step 2-1-2. If not, increase the stake of the updated validator by z3, delete the identified node from the feature set S, and return to step 2-1-2. z1, z2, and z3 are three preset stake adjustment values; z1 and z3 are both greater than z2.

[0020] 2-2. A corresponding candidate recommendation set is formed for each validator; each validator selects several subgroup management nodes from its corresponding candidate recommendation set and votes as block producers.

[0021] 2-3. All validators who received votes are selected as candidates for block production; the score of each candidate for block production is calculated based on the voting results of all validators and the size of the equity of each validator. The score of a candidate for block production is equal to the sum of the equity of all validators who voted for the candidate for block production; the candidates for block production with the largest scores are selected as the final block producers.

[0022] Step 3: All operations of nodes in the IoT layer generate corresponding Ethereum transactions; each block producer takes turns to package several Ethereum transactions into device blocks, which enter the blockchain after consensus.

[0023] Preferably, in step 1, the physical identifiers of different nodes are processed separately using hash technology and digital signature technology to form unique digital identifiers.

[0024] Preferably, the digital identification is defined as a device structure of a node; the device structure includes a device address, a device identifier, a subgroup management node address where the device is located, a metadata information hash, a firmware information hash, a timestamp, and a digital signature. The device structure can generate a device information block in the blockchain.

[0025] Preferably, the working cycle is greater than or equal to 12 hours.

[0026] Preferably, the value of z1 is 10; the value of z2 is 1 and the value of z3 is 10.

[0027] Preferably, the node operations described in step 3 include registration, information update and query.

[0028] Preferably, the consensus condition described in step 3 is that it is verified by greater than or equal to 2 / 3 of the validators.

[0029] Preferably, the device block in step 3 includes a block header and a block body, wherein the block header includes a parent block hash value, a Merkle tree root node, a timestamp, and a block producer list. The block body includes the hash values ​​of all transactions packaged in the block.

[0030] Preferably, the candidate recommendation set of all validators in step 2-2 is provided by the block producer voting algorithm; the specific process of the block producer voting algorithm is as follows:

[0031] First, the Q table is initialized according to the voting behavior of the validator in the previous work cycle, the set of elected block producers and the block production results of each block producer; then, the ε-greedy strategy is used to select the voting behavior with the greatest benefit; after executing the selected voting behavior, enter the new state and obtain an instantaneous reward; then use the maximum Q value and reward in the new state to update the Q value of the previous state, perform several iterations until the maximum number of iterations is reached or convergence, complete the Q table update, and output the candidate recommendation set.

[0032] In a second aspect, the present invention provides a blockchain-based Internet of Things system based on a lightweight consensus algorithm, which is used to execute the aforementioned Internet of Things method, and includes a cloud server layer and an Internet of Things layer. The Internet of Things layer is divided into multiple Internet of Things subgroups. Each Internet of Things subgroup consists of a management node and multiple subgroup ordinary nodes. Each Internet of Things subgroup is controlled by a control system. The control module includes a device management system, an underlying blockchain system and a client. The device management system includes a subgroup management node manager, a subgroup ordinary node manager, a data transmission module and a communication network management module. The underlying blockchain system is used to realize the functions of IoT terminal information storage, identity authentication and interactive evidence management.

[0033] The present invention has the following beneficial effects:

[0034] 1. The present invention realizes the trusted management of the Internet of Things based on blockchain technology, reduces the single point of failure and network performance bottleneck of centralized trust, and improves the scalability of the system. Blockchain provides reliable distributed data authentication and evidence storage for Internet of Things devices. Participants in the system can verify the authenticity of devices and data at any time. Blockchain ensures that the data generated in the Internet of Things system is open, transparent and cannot be tampered with, and the node behavior is traceable and undeniable.

[0035] 2. The present invention establishes a two-layer architecture including a cloud server layer and an IoT layer, which helps to realize offline storage and in-depth analysis of long-term large-scale data while ensuring low latency of real-time interaction of the IoT. At the same time, the IoT layer realizes privacy protection of sensitive data by dividing subgroups, reducing global management energy consumption.

[0036] 3. The present invention adopts Q-Learning to improve the equity proof consensus algorithm to achieve lightweight and low-energy blockchain consensus, which can effectively shorten the energy consumption and delay of blockchain consensus. At the same time, it updates the rights of validators through consensus and block production results, further enhancing the reliability of the system.

[0037] 4. The present invention has developed a blockchain IoT identity management system based on Ethereum, which realizes the identity authentication management of IoT entities. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a schematic diagram of the present invention.

[0039] Figure 2 It is a functional module diagram of the present invention.

[0040] Figure 3 It is a schematic diagram of the visual interface of the client in the present invention.

[0041] Figure 4 It is the working flow chart of the equipment management system in the present invention.

[0042] Figure 5 It is a schematic diagram of the device structure in the present invention.

[0043] Figure 6 It is a structural diagram of the device block in the present invention.

[0044] Figure 7 It is a schematic diagram of the working process of the present invention.

[0045] Figure 8 A flow chart for updating validator stake in each work cycle of the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with the accompanying drawings.

[0047] like Figure 1 As shown in the figure, a blockchain-based IoT system based on a lightweight consensus algorithm is divided into two layers: cloud server layer 1 and IoT layer 2. The upper cloud server layer 1 is responsible for data backup, background management, and data analysis and mining. The lower IoT layer 2 is divided into multiple IoT subgroups. Each IoT subgroup consists of a management node and multiple subgroup ordinary nodes. The subgroup management node is an IoT subgroup management node device with high computing power, such as a smart phone, a portable computer, etc., which is responsible for managing the subgroup and maintaining the underlying blockchain system; while the subgroup ordinary nodes are resource-constrained IoT devices, such as cameras, sensors, etc., which are deployed in the target area to collect or transmit data.

[0048] The functions of the subgroup management node include general management functions and blockchain functions. Among them, the general management function includes registering with the platform, managing the subgroup's equipment and data information, processing and broadcasting the subgroup's data to other subgroups. The blockchain function includes: assuming the role of a validator in the IoT control method, participating in consensus, participating in voting to elect block producers, and the subgroup management node may also be elected as a block producer to participate in block production.

[0049] The plane formed by the management nodes of different subgroups can be regarded as the management resource sublayer, while the communication network between device nodes in the same IoT subgroup can be regarded as the device sublayer. Devices within a subgroup can request and send data to each other to provide alliance services. For example, in a smart home environment, when the owner enters the door, the sound or video sensor will issue a command to control the light sensor to turn on the light, the temperature sensor to turn on the air conditioner, etc., to achieve intelligent linkage services. However, when IoT devices interact with data and behaviors, there is often a risk of identity fraud or data exposure. For example, tampering or forging smart keys to achieve vehicle theft, and stealing private data through forged data receiving devices. Therefore, identity authentication and secure and trusted management are key links in IoT interaction.

[0050] The IoT system is divided into trust domains. The trust relationship within the same IoT subgroup is called intra-domain trust, and the trust relationship between different IoT subgroups is called inter-domain trust. The underlying blockchain system mainly stores and authenticates the information of IoT devices, and uses hashing technology and digital signature technology to process the physical identification of the device to form a unique digital identification.

[0051] like Figure 2 As shown in the figure, each IoT subgroup is controlled by a control system. At the same time, the control system can aggregate the data uploaded by each IoT subgroup; the control system can be installed on a cloud server or on a central console in the IoT layer 2. The control module includes a device management system, an underlying blockchain system, and a client. The client includes a Web front-end module for subgroup management node login, subgroup ordinary node registration, information authentication, modification, and system visualization management. Figure 3 As shown in the figure, the client implements a visual interface, through which devices registered in the device management system can be managed. For example, the subgroup management node can vote to elect a block producer by clicking the device icon.

[0052] The equipment management system includes a subgroup management node manager, a subgroup common node manager, a data transmission module and a communication network management module.

[0053] The underlying blockchain system is used to implement the functions of IoT terminal information storage, identity authentication and interactive evidence management. The underlying blockchain system generates blocks and stores information and evidence through voting devices.

[0054] The workflow of the equipment management system is as follows: Figure 4 As shown in Figure 2. The functions of the cloud server include: receiving and verifying platform registration information, saving a copy of the blockchain ledger, and implementing offline analysis and data mining of interaction data. The functions of the IoT platform include: registering with the cloud server through the communication network, receiving and verifying the interaction data delivered by the subgroup through the blockchain, and uploading the authenticated interaction data to the cloud.

[0055] The functions of ordinary nodes in a subgroup include: registering with the subgroup management node, reading, transmitting and sharing local data.

[0056] Both subgroup management nodes and subgroup common nodes have the functions of registration, device information editing and search.

[0057] The working method of the blockchain Internet of Things system includes the following steps:

[0058] Step 1: Register the subgroup management node and subgroup common node to the blockchain IoT system. Use hashing technology and digital signature technology to process the physical identification of different nodes separately to form a unique digital identification. The specific process of forming a digital identification is as follows: Figure 5 As shown in the figure, specifically: define the device structure of all nodes separately; the device structure generates a device information block in the blockchain through a smart contract. The device structure contains the device address, device identifier, the address of the subgroup management node where the device is located, metadata information hash, firmware information hash, timestamp and digital signature.

[0059] After completing the registration, all node operations will be published to the blockchain system in the form of transactions. Among them, subgroup ordinary nodes also have the function of transferring ownership. If the device is successfully found, the device related information and operation records will be echoed, and if the search fails, the failure information will be echoed.

[0060] Step 2: Figure 7 and 8 As shown in the figure, each node in the IoT system is divided into three roles: ordinary members (devices), validators, and block producers. The validator is the subgroup management node; the validator votes to elect the block producer to produce blocks. In the initial state, each validator has the same rights and interests (expressed in numerical form).

[0061] Each work cycle is used to update the rights of the validator and re-elect the block producer. In this embodiment, the work cycle is 24 hours. The process of updating the rights of the validator and re-electing the block producer is as follows:

[0062] 2-1. Update the rights and interests of all validators. The specific process is as follows:

[0063] 2-1-1. Take the equity of the updated validator in the previous work cycle as the initial equity. Take the subgroup management nodes recommended as block producers by the updated validator in the previous work cycle as the feature set S.

[0064] 2-1-2. Determine whether the feature set S is an empty set; if so, it means that the validator's rights update is completed. If not, go to step 2-1-3.

[0065] 2-1-3. Take a subgroup management node in the feature set S as the identified node; determine whether the identified node becomes a block producer; if so, proceed to step 2-1-4; if not, delete the identified node from the feature set S and return to step 2-1-2.

[0066] 2-1-4. Determine whether the identified node has become an honest node; if so, proceed to step 2-1-5; if not, reduce the stake of the updated validator by 10, delete the identified node from the feature set S, and return to step 2-1-2.

[0067] 2-1-5. Determine whether the identified node is the updated validator itself. If so, increase the stake of the updated validator by 10, delete the identified node from the feature set S, and return to step 2-1-2. If not, increase the stake of the updated validator by 1, delete the identified node from the feature set S, and return to step 2-1-2.

[0068] According to the loop operation of steps 2-1-2 to 2-1-5, the nodes in the feature set S gradually decrease until the feature set S becomes an empty set, and the rights and interests of the updated validator are updated.

[0069] 2-2. The block producer voting algorithm based on Q-Learning is used to recommend block producers to validators, and a corresponding candidate recommendation set is formed for each validator. Each validator selects several subgroup management nodes from its corresponding candidate recommendation set and votes as block producers.

[0070] The specific process of the Q-Learning-based block producer voting algorithm is as follows: First, the Q table is initialized according to the voting behavior of the validator in the previous work cycle, the set of elected block producers and the block production results of each block producer; then, the ε-greedy strategy is used to select the voting behavior with the greatest benefit; after executing the selected voting behavior, enter the new state and obtain an instantaneous reward; then use the maximum Q value and reward in the new state to update the Q value of the previous state, perform several iterations until the maximum number of iterations is reached or convergence, complete the Q table update, and output the candidate recommendation set.

[0071] The expression for Q value update is shown in Equation 1:

[0072] Q(s,a)=Q(s,a)+α[r+γmax(Q(s',a')-Q(s,a))] Formula 1

[0073] Where s represents the state, a represents the behavior, Q(s,a) is the state-active Q-value function, r represents the profit at time t, s' represents the state at the next moment, a' represents the best behavior selected under the s' state, γ represents the discount factor, α represents the learning rate, γ, α∈(0,1], and max(Q(s',a')-Q(s,a)) represents the incremental profit from entering state s' from state s.

[0074] The Q-Learning-based block producer voting algorithm adopts a greedy strategy in each step of behavior selection. After several iterations, Q(s,a) will converge to the optimal solution π * (s,a)=argmax a Q(s,a). π * represents the optimal strategy in state s.

[0075] 2-3. All validators who have received votes are selected as candidates for block production; the block producer score of each candidate is calculated based on the voting results of all validators and the equity of each validator. The block producer score of a candidate is equal to the sum of the equity of all validators who voted for the candidate; the candidates with the largest block producer scores are selected as the final block producers.

[0076] Step 3: All operations of nodes in IoT layer 2 generate corresponding Ethereum transactions; each block producer takes turns to package several Ethereum transactions into device blocks, which enter the blockchain after consensus. Node operations include registration, information update and query. The consensus condition is that more than or equal to 2 / 3 of the validators have passed the verification.

[0077] like Figure 6 As shown, the device block includes a block header and a block body; the block header contains information including the parent block hash value, Merkle tree root node, timestamp, and block producer list. The block body contains the hash values ​​of all transactions packaged in this block.

[0078] Based on the above solution, the security of the Internet of Things system can be effectively improved.

Claims

1. A blockchain-based Internet of Things method based on a lightweight consensus algorithm, characterized by: The adopted Internet of Things system comprises a cloud server layer (1) and an Internet of Things layer (2); the Internet of Things layer (2) is divided into a plurality of Internet of Things subgroups; each Internet of Things subgroup is composed of a management node and a plurality of subgroup common nodes; The IoT method includes the following steps: Step 1: Register the subgroup management node and subgroup common nodes to the IoT system; Step 2: Set three roles: ordinary members, validators, and block producers. The validator is the subgroup management node; the ordinary member is the subgroup ordinary node; the block producer is elected by the validator; in the initial state, each validator has the same rights and interests; Each work cycle updates the rights of validators and re-elects block producers. The process is as follows: 2-1. Update the rights and interests of all validators; 2-1-1. Take the equity of the updated validator in the previous work cycle as the initial equity; take the subgroup management nodes recommended as block producers by the updated validator in the previous work cycle as the feature set S; 2-1-2. Determine whether the feature set S is an empty set; if so, it means that the validator's rights update is completed and go to step 2-2; if not, go to step 2-1-3; 2-1-3. Take a validator in the feature set S as the identified node; determine whether the identified node becomes a block producer; if so, proceed to step 2-1-4; if not, delete the identified node from the feature set S and return to step 2-1-2; 2-1-4. Determine whether the identified node has become an honest node; if so, proceed to step 2-1-5; if not, reduce the stake of the updated validator by z1; and delete the identified node from the feature set S, and return to step 2-1-2; 2-1-5. Determine whether the identified node is the updated validator itself. If so, increase the stake of the updated validator by z2, delete the identified node from the feature set S, and return to step 2-1-2; if not, increase the stake of the updated validator by z3, delete the identified node from the feature set S, and return to step 2-1-2; z1, z2 and z3 are three preset stake adjustment values; z1 and z3 are both greater than z2; 2-2. A corresponding candidate recommendation set is formed for each validator; each validator selects several subgroup management nodes from its corresponding candidate recommendation set and votes as the block producer; The candidate recommendation set for all validators is provided by the block producer voting algorithm; the specific process of the block producer voting algorithm is as follows: First, the Q table is initialized according to the voting behavior of the validator in the previous work cycle, the set of elected block producers and the block production results of each block producer; then, the ε-greedy strategy is used to select the voting behavior with the greatest benefit; after executing the selected voting behavior, enter the new state and obtain an instantaneous reward; then use the maximum Q value and reward in the new state to update the Q value of the previous state, perform several iterations until the maximum number of iterations is reached or convergence is achieved, complete the Q table update, and output the candidate recommendation set; The expression for Q value update is shown in Equation 1: Q(s,a)=Q(s,a)+α[r+γmax(Q(s',a')-Q(s,a))] Formula 1 Where s represents the state, a represents the behavior, Q(s,a) is the state-active Q-value function, r represents the benefit at time t, s' represents the state at the next moment, a' represents the best behavior selected under the s' state, γ represents the discount factor, α represents the learning rate, γ, α∈(0,1], max(Q(s',a')-Q(s,a)) represents the incremental benefit from entering state s' from state s; The Q-Learning-based block producer voting algorithm adopts a greedy strategy in each step of behavior selection. After several iterations, Q(s,a) will converge to the optimal solution π * (s,a)=argmax a Q(s,a);π * represents the optimal strategy in state s; 2-3. All validators who have received votes are selected as candidates for block producers. The block producer score of each candidate is calculated based on the voting results of all validators and the equity of each validator. The block producer score of a candidate is equal to the sum of the equity of all validators who voted for the candidate. Several candidates with the largest block producer scores are selected as the final block producers. Step 3: All operations of the nodes in the IoT layer (2) generate corresponding Ethereum transactions; each block producer takes turns to package several Ethereum transactions into device blocks, which enter the blockchain after consensus.

2. According to claim 1, a blockchain architecture Internet of Things method based on a lightweight consensus algorithm is characterized by: In step 1, the physical identifiers of different nodes are processed separately using hash technology and digital signature technology to form unique digital identifiers.

3. According to claim 2, a blockchain architecture Internet of Things method based on a lightweight consensus algorithm is characterized in that: The digital identification will be loaded into the device structure of the node; the device structure includes the device address, device identifier, the address of the subgroup management node where the device is located, metadata information hash, firmware information hash, timestamp and digital signature; the device structure can generate a device information block in the blockchain.

4. The method for Internet of Things based on blockchain architecture and lightweight consensus algorithm according to claim 1, characterized in that: The working cycle is greater than or equal to 12 hours.

5. The method for Internet of Things based on blockchain architecture and lightweight consensus algorithm according to claim 1, characterized in that: The value of z1 is 10; the value of z2 is 1 and the value of z3 is 10.

6. The method for Internet of Things based on blockchain architecture and lightweight consensus algorithm according to claim 1, characterized in that: The operations of the node described in step 3 include registration, information update and query.

7. The method for Internet of Things based on blockchain architecture and lightweight consensus algorithm according to claim 1, characterized in that: The consensus condition described in step 3 is that greater than or equal to 2 / 3 of the validators have verified it.

8. The method for Internet of Things based on blockchain architecture and lightweight consensus algorithm according to claim 1, characterized in that: The device block described in step 3 includes a block header and a block body; wherein the block header includes the parent block hash value, Merkle tree root node, timestamp, and block producer list; and the block body includes the hash values ​​of all transactions packaged in this block.

9. A blockchain-based Internet of Things system based on a lightweight consensus algorithm, characterized by: Used to execute the Internet of Things method as described in any one of claims 1 to 8, it includes a cloud server layer (1) and an Internet of Things layer (2); the Internet of Things layer (2) is divided into multiple Internet of Things subgroups; each Internet of Things subgroup consists of a management node and multiple subgroup ordinary nodes; each Internet of Things subgroup is controlled by a control system; the control module includes a device management system, an underlying blockchain system and a client; the device management system includes a subgroup management node manager, a subgroup ordinary node manager, a data transmission module and a communication network management module; the underlying blockchain system is used to realize the functions of IoT terminal information storage, identity authentication and interactive evidence management.

Images