Block chain computing power-based distributed virtual platform building method and system

Abstract

The invention discloses a distributed virtual platform building method and system based on block chain computing power, and belongs to the technical field of virtual platforms. The invention discloses a block chain-based computing power distributed virtual platform building method. The method comprises the following steps of S1, creating a physical computing power resource pool; s2, constructing a chain type block structure; s3, obtaining a block chain computing power distributed platform; and S4, obtaining a block chain computing power distributed virtual platform. And S5, tracking the utilization rate and the network bandwidth index of the block chain computing power distributed virtual platform in real time. According to the method, the problem that the whole system may be paralyzed once a core node has a fault or is attacked due to dependence on centralized server management data and computing power in the prior art is solved, the virtual platform is established by utilizing the block chain distributed platform, so that the virtual platform does not need to depend on a single central node, and the system reliability is improved. All computing power resources are jointly maintained and verified by a plurality of nodes in the network, so that the stability and the disaster recovery capability are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of virtual platform technology, and specifically to a method and system for building a distributed virtual platform based on blockchain computing power. Background Art

[0002] Virtualization refers to the process of running a user's desktop within a virtual machine residing on a server. Server-based desktop virtualization systems can provide each user with a personalized desktop while also ensuring centralized management and security. The server in such a system may include storage for virtual desktop images and system configuration information, as well as software components that provide virtual machines and allow users to connect to them.

[0003] A Chinese patent with publication number CN115908763A discloses a method for building a virtual visualization intelligent manufacturing platform, which includes the following steps: Step 1: The physical layer constructs a physical intelligent manufacturing platform; Step 2: The perception layer collects static and dynamic data of the platform and transmits the collected data to the data layer, which stores, processes and analyzes the data; Step 3: At the same time, the virtual layer performs three-dimensional modeling of the equipment, builds the platform scene, and sets up the virtual platform environment; Step 4: The application layer develops the interaction logic between the user and the virtual platform, designs the human-computer interaction interface, displays, updates, and operates information, and publishes and debugs the system client of the intelligent manufacturing platform. By using three-dimensional virtual visualization technology, real-time visual tracking of the production process is realized, information exchange between the management layer and the execution layer is opened up, and the production efficiency of the intelligent manufacturing system is improved.

[0004] In actual use, the above-mentioned patent relies on centralized servers to manage data and computing power. Once the core node fails or is attacked, the entire system may be paralyzed. Therefore, it does not meet existing needs. In response to this, we propose a method and system for building a distributed virtual platform based on blockchain computing power. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for building a distributed virtual platform based on blockchain computing power, thereby solving the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for building a distributed virtual platform based on blockchain computing power, comprising the following steps:

[0007] S1: Create a physical computing resource pool on the Hypervisor virtual platform and allocate computing resources to form a physical node group;

[0008] S2: Construct a chain block structure and use the Merkle tree to optimize data verification to verify the efficiency of the chain block structure;

[0009] S3: Plan the nodes on the chain block structure and deploy the computing power of the nodes to obtain a distributed blockchain computing power platform;

[0010] S4: Connect the physical computing resource pool with the blockchain computing power distributed platform, encapsulate the physical nodes on the physical computing resource pool as callable resources on the blockchain computing power distributed platform, and obtain the blockchain computing power distributed virtual platform;

[0011] S5: Track the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time, and use Apache Mesos to distribute tasks across nodes based on the tracking results.

[0012] Preferably, the chain block structure specifically includes:

[0013] The data layer is used to design chain storage structures and optimize block verification efficiency;

[0014] The network layer is used to divide the dedicated communication channels between nodes and physical node groups on the blockchain computing power distribution platform through VLANs, assign fixed IP addresses to nodes, and combine dynamic domain name resolution to ensure the accessibility of nodes across regions;

[0015] The consensus layer is used to select a consensus mechanism that suits the computing power requirements and balance computing power consumption and efficiency;

[0016] The incentive layer is used to provide economic rewards to encourage more nodes to participate in the security verification of the blockchain;

[0017] Dynamic resource allocation layer, used to implement priority scheduling and elastic expansion of computing tasks using smart contract algorithms;

[0018] Preferably, the planning of nodes on the chain block structure specifically includes:

[0019] Divide into full nodes, light nodes and computing nodes. Full nodes are used to store complete data, light nodes are used to verify key data, and computing nodes are used for complex calculations, while optimizing the division of computing power.

[0020] Set the data storage path, synchronization mode, and network connection parameters of the full node based on the geographical location and bandwidth conditions;

[0021] Set up an authentication mechanism for light nodes, assign operation permissions to different roles, and automatically download and verify data on the chain block structure after starting the node.

[0022] Preferably, the use of Merkle tree to optimize the efficiency of data verification chain block structure specifically includes:

[0023] All transaction data in the chain block structure are recursively hashed through the Merkle tree to generate a unique root hash value and store it in the block header;

[0024] When verifying a single transaction, the path node from the transaction hash to the root hash is provided, and verification is completed through layer-by-layer hash calculation and comparison;

[0025] When adding a new transaction, the affected branch nodes are updated by pre-filling the default hash value. During verification, the hash calculations of different subtree layers can be performed in a distributed manner, utilizing multi-core processors to accelerate the tree construction process.

[0026] The system for building a distributed virtual platform based on blockchain computing power is applied in a method for building a distributed virtual platform based on blockchain computing power, including:

[0027] Create a module for deploying high-performance server clusters, distributed storage systems, and high-speed network equipment on the hypervisor virtual platform to form a physical computing resource pool;

[0028] The computing power distribution module is used to allocate independent computing power resources to each node in the physical computing power resource pool and deploy the computing power resources;

[0029] Chain block structure, used to design chain storage structure, establish dedicated communication channels, and use smart contract algorithms to achieve priority scheduling and elastic expansion of computing tasks;

[0030] The platform building module is used to encapsulate the physical nodes on the physical computing power resource pool into callable resources on the blockchain computing power distributed platform to obtain the blockchain computing power distributed virtual platform;

[0031] An integrated monitoring module is used to track the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time, and use Apache Mesos to distribute tasks across nodes based on the tracking results.

[0032] Preferably, the computing power distribution module includes:

[0033] The computing power scheduling unit is used to allocate computing power resources on the corresponding nodes to each task based on the service priority queue of the physical computing power resource pool and the computing power available resource pool registry;

[0034] The computing power deployment unit is used to deploy the computing power of multiple computing power resources by utilizing the communication reference information between the physical computing power resource pools and the computing tasks of the blockchain computing power distribution platform.

[0035] Preferably, the computing power deployment unit includes:

[0036] Integrate the general computing, accelerated computing, and edge computing resources of the blockchain computing power distributed platform to obtain computing tasks of the integrated blockchain computing power distributed platform;

[0037] Based on the communication reference information between physical computing resource pools and the computing tasks of the blockchain computing power distributed platform, Ansible is used to deploy core nodes in the central cloud, connect computing nodes to the edge cloud, and connect edge nodes to end-side devices;

[0038] The central cloud is responsible for global resource scheduling and complex model training, the edge cloud is used for data cleaning and lightweight inference, and the end-side devices are used to collect real-time data to obtain the computing power deployment results of multiple computing resources in the physical computing resource pool;

[0039] A physical node group is formed by utilizing the computing power deployment results of multiple computing resources in the physical computing power resource pool.

[0040] Preferably, the computing power scheduling unit includes:

[0041] Select the corresponding nodes for the tasks in the order of the tasks in the physical computing resource pool service priority queue;

[0042] Using a periodic step-by-step scheduling approach, high-value nodes are prioritized, and service orchestration action decisions are made for the selected high-value nodes to determine the computing power resources allocated to each task on the corresponding node.

[0043] When the computing power resources on any node are greater than or equal to the preset computing power resource threshold, any node is determined to be a high-value node.

[0044] Preferably, dynamically adjusting the computing power resource threshold using the computing power capacity corresponding to all nodes includes:

[0045] Real-time monitoring of the computing power fluctuation rate corresponding to each node;

[0046] Comparing the computing power volatility with a preset computing power volatility reference value;

[0047] When the computing power volatility exceeds the preset computing power volatility parameter value, the node corresponding to the computing power volatility exceeding the preset computing power volatility parameter value is selected as the selected node;

[0048] Performing difference processing on the computing power volatility corresponding to the selected node and the computing power volatility parameter value to obtain computing power volatility difference data between the computing power volatility and the computing power volatility parameter value;

[0049] Ratio processing is performed on the computing power fluctuation rate difference data corresponding to the selected node and its corresponding computing power fluctuation rate to obtain the computing power fluctuation ratio corresponding to the selected node;

[0050] Retrieve the task urgency index corresponding to the selected node; wherein the task urgency index is calculated as follows:

[0051] Task urgency index = number of overdue tasks / total number of processed tasks

[0052] Comparing the task urgency index corresponding to the selected node with its corresponding computing power fluctuation ratio;

[0053] When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the computing power resource threshold is adjusted.

[0054] Preferably, when there is a node whose task urgency index exceeds the task urgency index, the computing power resource threshold is adjusted, including:

[0055] When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the node whose task urgency index exceeds the computing power fluctuation ratio will be used as the target node;

[0056] The task backlog and utilization gradient values of the target node are retrieved. The task backlog and utilization gradient values corresponding to the target node are obtained using the following formula:

[0057] When the number of the target node is 1, the task backlog and utilization gradient values corresponding to the target node are used as gradient reference values;

[0058] When there are multiple target nodes, the standard deviation of the task backlog and utilization gradient values corresponding to the multiple target nodes is used as the gradient reference value;

[0059] The computing power resource threshold is adjusted using the gradient reference value combined with the spatiotemporal attenuation compensation coefficient corresponding to the target node.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] The present invention uses a blockchain distributed platform to build a virtual platform, so that the virtual platform does not need to rely on a single central node. All computing resources are jointly maintained and verified by multiple nodes in the network, avoiding system crashes caused by server failures or attacks on traditional centralized platforms, significantly improving stability and disaster recovery capabilities, ensuring business continuity and data integrity, and all the allocation, use and transaction processes of computing resources are recorded on the blockchain. Participants can query historical data in real time, realize transparent supervision of the entire process, effectively prevent resource abuse or fraud, significantly improve efficiency and reduce operating costs. Distributed computing power integrates idle resources. At the same time, the decentralized structure eliminates the intermediary fees of traditional cloud services, so that both resource providers and users can reduce costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a schematic diagram of the method for building a distributed virtual platform based on blockchain computing power of the present invention;

[0063] Figure 2 This is a schematic diagram of the system for building a distributed virtual platform based on blockchain computing power in the present invention. DETAILED DESCRIPTION

[0064] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0065] In order to solve the problem that the existing technology relies on centralized servers to manage data and computing power in actual use, once the core node fails or is attacked, it may cause the entire system to be paralyzed, please refer to Figure 1-Figure 2 , this embodiment provides the following technical solutions:

[0066] The method for building a distributed virtual platform based on blockchain computing power includes the following steps:

[0067] S1: Creates a physical computing resource pool on the Hypervisor virtual platform, allocates computing resources to form a physical node group, and supports dynamic expansion of CPU core count and memory size.

[0068] S2: Build a chain block structure and use the Merkle tree to optimize the efficiency of data verification chain block structure, achieving data redundancy and fast retrieval;

[0069] S3: Plan the nodes on the chain block structure and deploy the computing power of the nodes. Use high-computing-power nodes for transaction verification, and low-computing-power nodes for data synchronization. This will create a distributed blockchain computing power platform and achieve hierarchical resource utilization.

[0070] S4: Connect the physical computing resource pool with the blockchain computing power distributed platform, encapsulate the physical nodes on the physical computing resource pool as callable resources on the blockchain computing power distributed platform, and obtain the blockchain computing power distributed virtual platform;

[0071] S5: Track the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time, and use Apache Mesos to distribute tasks across nodes based on the tracking results.

[0072] Utilizing a distributed blockchain platform to build a virtual platform eliminates the need for a single central node. Instead, all computing resources are maintained and verified by multiple nodes across the network. This avoids system crashes caused by server failures or attacks, as is the case with traditional centralized platforms, significantly improving stability and disaster recovery. Once on-chain, computing power transaction records and task allocation data on the platform cannot be tampered with. Even if some nodes are attacked, the overall data can be restored through other nodes, ensuring business continuity and data integrity. All computing power resource allocation, usage, and transaction processes are recorded on the blockchain, allowing participants to access historical data in real time, achieving transparent oversight throughout the entire process and effectively preventing resource misuse and fraud. Processes such as computing power scheduling and fee settlement can be automated through pre-set smart contracts, reducing manual intervention. For example, contracts can dynamically match the optimal computing power node based on task requirements and trigger payments based on completion, significantly improving efficiency and reducing operational costs. Distributed computing power consolidates idle resources, enabling global scheduling through the blockchain platform and avoiding duplicate investment. Furthermore, the decentralized structure eliminates the intermediary fees of traditional cloud services, reducing costs for both resource providers and users.

[0073] Chain block structure, specifically including:

[0074] The data layer is used to design chain storage structures, such as timestamps and hash functions, and optimize block verification efficiency;

[0075] The network layer is used to divide the dedicated communication channels between nodes and physical node groups on the blockchain computing power distribution platform through VLANs, assign fixed IP addresses to nodes, and combine dynamic domain name resolution to ensure the accessibility of nodes across regions, thereby achieving efficient communication between nodes and dynamic scheduling of computing power;

[0076] The consensus layer is used to select a consensus mechanism that suits the computing power requirements, such as PoW, PoS, and DPoS, to balance computing power consumption and efficiency;

[0077] The incentive layer is used to provide economic rewards to encourage more nodes to participate in the security verification of the blockchain;

[0078] Dynamic resource allocation layer, used to implement priority scheduling and elastic expansion of computing tasks using smart contract algorithms;

[0079] Planning the nodes on the chain block structure, including:

[0080] Divide into full nodes, light nodes and computing nodes. Full nodes are used to store complete data, light nodes are used to verify key data, and computing nodes are used for complex calculations, while optimizing the division of computing power.

[0081] Set the data storage path, synchronization mode, and network connection parameters of the full node based on the geographical location and bandwidth conditions;

[0082] Set up an authentication mechanism for light nodes, assign operation permissions to different roles, and automatically download and verify data on the chain block structure after starting the node.

[0083] Merkle trees are used to optimize the efficiency of data verification chain block structures, including:

[0084] All transaction data within the chained block structure is recursively hashed through the Merkle tree to generate a unique root hash value that is stored in the block header. During verification, only the root hash needs to be compared to determine whether the entire data set has been tampered with, without the need to verify the original data one by one.

[0085] Leaf nodes store transaction hashes, and non-leaf nodes are generated by concatenating the hashes of child nodes and performing a secondary hash, which can reduce the complexity of large-scale data integrity verification.

[0086] When verifying a single transaction, the path node from the transaction hash to the root hash is provided, and verification is completed through layer-by-layer hash calculation and comparison;

[0087] Nodes do not need to download the entire block data. Instead, verification can be completed through the Merkle Root in the block header and a small amount of path hash, reducing network transmission by more than 90%;

[0088] When adding a new transaction, the affected branch nodes are updated by pre-filling the default hash value instead of reconstructing the entire tree structure. This is suitable for scenarios that require frequent insertion and deletion operations.

[0089] During verification, hash calculations at different subtree layers can be executed in a distributed manner, using multi-core processors to accelerate the tree construction process. In a sharded blockchain, each shard can independently generate a Merkle Root, and then implement cross-shard data verification through the upper-level tree structure.

[0090] The system for building a distributed virtual platform based on blockchain computing power is applied in a method for building a distributed virtual platform based on blockchain computing power, including:

[0091] Create a module for deploying high-performance server clusters, distributed storage systems, and high-speed network equipment on the hypervisor virtual platform to form a physical computing resource pool;

[0092] The computing power distribution module is used to allocate independent computing power resources to each node in the physical computing power resource pool and deploy the computing power resources;

[0093] The chain block structure is used to design a chain storage structure, establish a dedicated communication channel, and use smart contract algorithms to achieve priority scheduling and elastic expansion of computing tasks:

[0094] By building a chain block structure, the security and immutability of data are ensured. Each block contains a unique hash value, which is calculated based on all transaction data and other information within the chain block structure (such as the hash value of the previous block). If any information within the chain block structure is changed, even a small change, the hash value of the block will be completely changed, which will in turn affect all subsequent blocks, making tampering with data on the blockchain extremely difficult and costly. Each block is linked together by containing the hash value of the previous block to form a chain. Once the data is recorded on the blockchain, it cannot be modified or deleted, ensuring the consistency and credibility of the data and increasing the difficulty of tampering.

[0095] The platform building module is used to encapsulate the physical nodes on the physical computing power resource pool into callable resources on the blockchain computing power distributed platform to obtain the blockchain computing power distributed virtual platform.

[0096] An integrated monitoring module is used to track the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time. Based on the tracking results, Apache Mesos is used to distribute tasks across nodes. The computing power allocation strategy is dynamically adjusted through smart contracts. For example, the GPU node capacity is automatically expanded during peak hours to meet complex computing needs.

[0097] By tracking the utilization rate of the distributed virtual platform of blockchain computing power and network bandwidth indicators in real time, all nodes can view the computing power allocation and bandwidth usage records in real time to ensure transparent operations. Through encrypted hash value storage, any abnormal adjustments will be identified by the nodes to ensure data authenticity and historical traceability. Real-time monitoring of computing power utilization can identify high-load nodes, dynamically allocate tasks to avoid resource waste, and improve the overall efficiency of the platform. By monitoring network bandwidth indicators, data transmission paths can be adjusted in a timely manner to reduce congestion and increase transaction processing speed.

[0098] The computing power distribution module includes:

[0099] The computing power scheduling unit is used to allocate computing power resources on the corresponding nodes to each task based on the service priority queue of the physical computing power resource pool and the computing power available resource pool registry;

[0100] The computing power deployment unit is used to deploy the computing power of multiple computing power resources by utilizing the communication reference information between the physical computing power resource pools and the computing tasks of the blockchain computing power distribution platform.

[0101] The computing power deployment module can significantly improve the performance and efficiency of the computing power network by optimizing hardware resource configuration and software tuning. The computing power deployment module can support the construction and operation of ultra-large-scale intelligent computing centers, reduce external connections, improve data transmission efficiency, and thus reduce power consumption.

[0102] Computing power deployment unit, including:

[0103] Integrate the general computing, accelerated computing, and edge computing resources of the blockchain computing power distributed platform to obtain computing tasks of the integrated blockchain computing power distributed platform;

[0104] Based on the communication reference information between physical computing resource pools and the computing tasks of the blockchain computing power distributed platform, Ansible is used to deploy core nodes in the central cloud, connect computing nodes to the edge cloud, and connect edge nodes to end-side devices to achieve elastic expansion and automated configuration management. The communication reference information is the communication resources consumed by data transmission between physical computing resource pools;

[0105] The central cloud is responsible for global resource scheduling and complex model training, the edge cloud is used for data cleaning and lightweight inference, and the end-side devices are used to collect real-time data to obtain the computing power deployment results of multiple computing resources in the physical computing resource pool;

[0106] A physical node group is formed by utilizing the computing power deployment results of multiple computing resources in the physical computing power resource pool.

[0107] Computing power scheduling unit, including:

[0108] Select the corresponding nodes for the tasks in the order of the tasks in the physical computing resource pool service priority queue;

[0109] Using a periodic step-by-step scheduling approach, high-value nodes are prioritized, and service orchestration action decisions are made for the selected high-value nodes to determine the computing power resources allocated to each task on the corresponding node.

[0110] When the computing power resources on any node are greater than or equal to the preset computing power resource threshold, any node is determined to be a high-value node.

[0111] Specifically, dynamically adjusting the computing power resource threshold using the computing power capacity corresponding to all nodes includes:

[0112] Real-time monitoring of the computing power fluctuation rate corresponding to each node;

[0113] Comparing the computing power volatility with a preset computing power volatility reference value;

[0114] When the computing power volatility exceeds the preset computing power volatility parameter value, the node corresponding to the computing power volatility exceeding the preset computing power volatility parameter value is selected as the selected node;

[0115] Performing difference processing on the computing power volatility corresponding to the selected node and the computing power volatility parameter value to obtain computing power volatility difference data between the computing power volatility and the computing power volatility parameter value;

[0116] Ratio processing is performed on the computing power fluctuation rate difference data corresponding to the selected node and its corresponding computing power fluctuation rate to obtain the computing power fluctuation ratio corresponding to the selected node;

[0117] Retrieve the task urgency index corresponding to the selected node; wherein the task urgency index is calculated as follows:

[0118] Task urgency index = number of overdue tasks / total number of processed tasks

[0119] Comparing the task urgency index corresponding to the selected node with its corresponding computing power fluctuation ratio;

[0120] When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the computing power resource threshold is adjusted.

[0121] The technical solution described above provides the following benefits: By monitoring the computing power fluctuation rate of each node in real time and comparing it with a preset reference value, changes in node computing power can be captured promptly. When the computing power fluctuation rate exceeds a preset parameter, a series of computational analyses are performed on the relevant nodes, such as calculating the computing power fluctuation ratio. This enables the system to flexibly identify nodes with abnormal computing power fluctuations based on the dynamic changes in node computing power, providing a basis for subsequent adjustments and enhancing the system's adaptability to dynamic changes in computing power resources. A task urgency index (number of timed tasks divided by total number of processed tasks) is introduced and compared with the computing power fluctuation ratio. The computing power resource threshold is adjusted when a node's task urgency index exceeds the computing power fluctuation ratio. This demonstrates that the system not only considers computing power fluctuations but also incorporates the urgency of task execution. This allows for more accurate dynamic adjustment of computing power resource thresholds based on the actual business needs and computing power status of the node, enabling the system to adapt effectively to diverse business scenarios and computing power fluctuations, ensuring stable system performance. Calculating the computing power fluctuation ratio for selected nodes, combined with the task urgency index, enables a more comprehensive and accurate assessment of node status. For nodes experiencing significant computing power fluctuations and high task urgency, timely adjustment of computing power resource thresholds can prevent task delays or resource waste caused by inappropriate computing power allocation. For example, if a node experiences significant computing power fluctuations but low task urgency, immediate resource adjustment may not be necessary. However, for nodes experiencing both high computing power fluctuations and low task urgency, adjusting the computing power resource threshold can prioritize task processing for these nodes, ensuring that computing power resource allocation is more aligned with the node's actual needs and improving resource allocation efficiency. This dynamic adjustment mechanism, based on node computing power and task status, balances computing power resource allocation across nodes, preventing idle computing power on some nodes and hindering task execution on others due to insufficient computing power. This improves overall computing power resource utilization and optimizes system performance. Monitoring and addressing computing power fluctuations can proactively identify abnormal node computing power fluctuations. By adjusting the computing power resource threshold, the risk of unstable node task processing or even system failures caused by excessive computing power fluctuations can be mitigated, ensuring system stability. Adjustments made in conjunction with the task urgency index can ensure that critical tasks are stably executed with appropriate computing power support, reduce problems such as task timeouts and failures caused by improper allocation of computing resources, and further improve the stability of the system in business processing.

[0122] Specifically, when there is a node whose task urgency index exceeds the task urgency index, the computing power resource threshold is adjusted, including:

[0123] When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the node whose task urgency index exceeds the computing power fluctuation ratio will be used as the target node;

[0124] The task backlog and utilization gradient values of the target node are retrieved. The task backlog and utilization gradient values corresponding to the target node are obtained using the following formula:

[0125]

[0126] Among them, J represents the task backlog and utilization gradient value corresponding to the target node; M represents the number of backlog tasks corresponding to the target node; N represents the total number of processing tasks corresponding to the target node;

[0127] When the number of the target node is 1, the task backlog and utilization gradient values corresponding to the target node are used as gradient reference values;

[0128] When there are multiple target nodes, the standard deviation of the task backlog and utilization gradient values corresponding to the multiple target nodes is used as the gradient reference value;

[0129] Adjusting the computing power resource threshold using the gradient reference value combined with the spatiotemporal attenuation compensation coefficient corresponding to the target node;

[0130] The adjusted computing power resource threshold is obtained through the following formula:

[0131]

[0132] Where D represents the computing power resource threshold after adjustment; D0 represents the computing power resource threshold before adjustment; ΔT represents the temporal and spatial attenuation compensation coefficient, which corrects the threshold deviation caused by the difference in computing power between the edge and the cloud; D m represents the preset maximum theoretical threshold; δ represents the gradient reference value.

[0133] Specifically, It reflects the backlog of tasks at the target node. The higher the proportion of backlog tasks to total tasks, the more serious the backlog. The meaning of P is not mentioned here, but from the formula, it is a further adjustment factor for the backlog. Adding 1 is to make the result reflect the correlation between task backlog and utilization on a certain basis. As a whole, a value J that can measure the gradient of the correlation between task backlog and utilization of target nodes is constructed. The larger the value, the more serious the task backlog is or the more resource utilization needs to be adjusted. The Sigmoid function is a commonly used activation function that maps input values to the range 0-1. Here, it is used to The calculation results are transformed nonlinearly so that the adjusted computing power resource threshold varies within a reasonable range to avoid unreasonable extreme values. D0 is the computing power resource threshold before adjustment, and ΔT is the spatiotemporal attenuation compensation coefficient. The sum of the two reflects the preliminary adjustment based on the original threshold taking into account the difference in computing power between the edge and cloud, and reflects the impact of different computing power environments on the threshold. δ is the gradient reference value, and 1+δ is used as the denominator to adjust the scaling degree of the numerator calculation result. The larger the gradient reference value, the more obvious the scaling effect on the numerator, reflecting the impact of task backlog and utilization gradient values on the adjustment of computing power resource thresholds. Overall principle: After the above calculation and Sigmoid function transformation, multiply it by the preset maximum theoretical threshold D m , the final adjusted computing power resource threshold D is obtained, which comprehensively considers the impact of the initial threshold, computing power environment differences, and node task load-related gradient values on the threshold.

[0134] The technical effect of the above technical solution is: by calculating the task backlog and utilization gradient value to measure the target node task backlog and resource utilization, it can accurately reflect the actual task load status of the node. When the number of target nodes is different, different methods are used to determine the gradient reference value (a single node directly uses its gradient value, and multiple nodes use the standard deviation), which can more reasonably integrate the information of multiple nodes, so that the subsequent adjustment of the computing power resource threshold based on this value is more accurate to meet the task requirements of different nodes. The introduction of the time-space attenuation compensation coefficient ΔT corrects the threshold deviation caused by the difference in computing power between the edge and the cloud, and takes into account the impact of the computing power difference on resource allocation in different computing environments, so that the computing power resource threshold adjustment can better adapt to complex computing environments and enhance the system's adaptability to different scenarios. The computing power resource threshold is adjusted jointly by the gradient reference value and the time-space attenuation compensation coefficient, and the differences in node task load and computing power environment are comprehensively considered to avoid excessive or insufficient allocation of computing power resources, make computing power resource allocation more reasonable, improve resource utilization efficiency, and ensure stable system operation. At the same time, This formula accurately calculates the task backlog and utilization gradient values, providing accurate basic data for the subsequent determination of the gradient reference value. Whether a single target node directly uses this value, or multiple target nodes calculate their standard deviation as the gradient reference value, they rely on this formula to accurately measure the node task situation, thereby ensuring the accuracy of the computing power resource threshold adjustment. In addition, By introducing ΔT to account for differences in computing power environments, the formula adapts to different computing scenarios. Using δ, combined with task backlogs and utilization gradient values, we rationally adjust computing resource thresholds for more balanced resource allocation. The Sigmoid function ensures that the adjusted thresholds remain within a reasonable range, ensuring stable system operation and improving performance indicators such as adaptability and rationality.

[0135] In summary, the present invention provides a method and system for building a distributed virtual platform based on blockchain computing power. This utilizes a distributed blockchain platform to build a virtual platform, eliminating the need for the virtual platform to rely on a single central node. Instead, all computing resources are maintained and verified by multiple nodes in the network. This avoids system crashes caused by server failures or attacks in traditional centralized platforms, significantly improving stability and disaster recovery. By constructing a chained block structure, data security and immutability are ensured. If any information within the chained block structure is altered, even a minor change, the hash value of the block will completely change, affecting all subsequent blocks. This makes tampering with data on the blockchain extremely difficult and costly. Each block is linked together by including the hash value of the previous block, forming a chain. Once data is recorded on the blockchain, it cannot be modified or deleted, ensuring data consistency and credibility while increasing the difficulty of tampering. This ensures transparent oversight of the entire process, effectively preventing resource abuse or fraud, significantly improving efficiency, and reducing operating costs. Distributed computing power integrates idle resources. At the same time, the decentralized structure eliminates the intermediary fees of traditional cloud services, reducing costs for both resource providers and users. Using the computing power deployment module to optimize hardware resource configuration and software tuning can significantly improve the performance and efficiency of the computing power network, support the construction and operation of ultra-large-scale intelligent computing centers, reduce external connections, improve data transmission efficiency, and thus reduce power consumption. By real-time monitoring of computing power utilization, high-load nodes can be identified, tasks can be dynamically allocated to avoid resource waste, and the overall efficiency of the platform can be improved. By monitoring network bandwidth indicators, data transmission paths can be adjusted in a timely manner to reduce congestion and increase transaction processing speed.

[0136] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0137] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for building a distributed virtual platform based on blockchain computing power, characterized in that: The following steps are involved: S1: Create a physical computing resource pool on the Hypervisor virtual platform and allocate computing resources to form a physical node group; S2: Construct a chain block structure and use the Merkle tree to optimize data verification to verify the efficiency of the chain block structure; S3: Plan the nodes on the chain block structure and deploy the computing power of the nodes to obtain a distributed blockchain computing power platform; S4: Connect the physical computing resource pool with the blockchain computing power distributed platform, encapsulate the physical nodes on the physical computing resource pool as callable resources on the blockchain computing power distributed platform, and obtain the blockchain computing power distributed virtual platform; S5: Tracks the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time, and uses Apache Mesos to distribute tasks across nodes based on the tracking results.

2. The method for building a distributed virtual platform based on blockchain computing power according to claim 1 is characterized in that: The chain block structure specifically includes: ‌Data layer, used to design chain storage structure and optimize block verification efficiency; The network layer is used to divide the dedicated communication channels between nodes and physical node groups on the blockchain computing power distribution platform through VLANs, assign fixed IP addresses to nodes, and combine dynamic domain name resolution to ensure the accessibility of nodes across regions; The consensus layer is used to select a consensus mechanism that suits the computing power requirements and balance computing power consumption and efficiency; The incentive layer is used to provide economic rewards to encourage more nodes to participate in the security verification of the blockchain; The dynamic resource allocation layer is used to implement priority scheduling and elastic expansion of computing tasks using smart contract algorithms.

3. The method for building a distributed virtual platform based on blockchain computing power according to claim 1 is characterized in that: The planning of nodes on the chain block structure specifically includes: Divide into full nodes, light nodes and computing nodes. Full nodes are used to store complete data, light nodes are used to verify key data, and computing nodes are used for complex calculations, while optimizing the division of computing power. Set the data storage path, synchronization mode, and network connection parameters of the full node based on the geographical location and bandwidth conditions; Set up an authentication mechanism for light nodes, assign operation permissions to different roles, and automatically download and verify data on the chain block structure after starting the node.

4. The method for building a distributed virtual platform based on blockchain computing power according to claim 1, characterized in that: The use of Merkle trees to optimize the efficiency of data verification chain block structures specifically includes: All transaction data in the chained block structure is recursively hashed through the Merkle tree to generate a unique root hash value and store it in the block header. When verifying a single transaction, the path node from the transaction hash to the root hash is provided, and verification is completed through layer-by-layer hash calculation and comparison; When adding a new transaction, the affected branch nodes are updated by pre-filling the default hash value. During verification, the hash calculations of different subtree layers can be performed in a distributed manner, utilizing multi-core processors to accelerate the tree construction process.

5. A system for building a distributed virtual platform based on blockchain computing power, applied in a method for building a distributed virtual platform based on blockchain computing power as claimed in claim 4, characterized in that: include: Create a module for deploying high-performance server clusters, distributed storage systems, and high-speed network equipment on the hypervisor virtual platform to form a physical computing resource pool; ‌ The computing power distribution module is used to allocate independent computing power resources to each node in the physical computing power resource pool and deploy the computing power resources; Chain block structure, used to design chain storage structure, establish dedicated communication channels, and use smart contract algorithms to achieve priority scheduling and elastic expansion of computing power tasks; The platform building module is used to encapsulate the physical nodes on the physical computing power resource pool into callable resources on the blockchain computing power distributed platform, thereby obtaining a blockchain computing power distributed virtual platform; An integrated monitoring module is used to track the utilization rate and network bandwidth indicators of the distributed virtual platform of blockchain computing power in real time, and use Apache Mesos to distribute tasks across nodes based on the tracking results.

6. The system for building a distributed virtual platform based on blockchain computing power according to claim 5 is characterized in that: The computing power distribution module includes: The computing power scheduling unit is used to allocate computing power resources on the corresponding nodes to each task based on the service priority queue of the physical computing power resource pool and the computing power available resource pool registry; The computing power deployment unit is used to deploy the computing power of multiple computing resources by utilizing the communication reference information between physical computing power resource pools and the computing tasks of the blockchain computing power distribution platform.

7. The system for building a distributed virtual platform based on blockchain computing power according to claim 6 is characterized in that: The computing power deployment unit includes: Integrate the general computing, accelerated computing, and edge computing resources of the blockchain computing power distributed platform to obtain computing tasks of the integrated blockchain computing power distributed platform; Based on the communication reference information between physical computing resource pools and the computing tasks of the blockchain computing power distributed platform, Ansible is used to deploy core nodes in the central cloud, connect computing nodes to the edge cloud, and connect edge nodes to end-side devices; The central cloud is responsible for global resource scheduling and complex model training, the edge cloud is used for data cleaning and lightweight inference, and the end-side devices are used to collect real-time data to obtain the computing power deployment results of multiple computing resources in the physical computing resource pool; A physical node group is formed by utilizing the computing power deployment results of multiple computing resources in the physical computing power resource pool.

8. The system for building a distributed virtual platform based on blockchain computing power according to claim 6 is characterized in that: The computing power scheduling unit includes: Select the corresponding nodes for the tasks in the order of the tasks in the physical computing resource pool service priority queue; Using a periodic step-by-step scheduling approach, high-value nodes are prioritized, and service orchestration action decisions are made for the selected high-value nodes to determine the computing power resources allocated to each task on the corresponding node. When the computing power resources on any node are greater than or equal to the preset computing power resource threshold, any node is determined to be a high-value node.

9. The system for building a distributed virtual platform based on blockchain computing power according to claim 8, characterized in that: Dynamically adjust the computing power resource threshold using the computing power capacity corresponding to all nodes, including: Real-time monitoring of the computing power fluctuation rate corresponding to each node; Comparing the computing power volatility with a preset computing power volatility reference value; When the computing power volatility exceeds the preset computing power volatility parameter value, the node corresponding to the computing power volatility exceeding the preset computing power volatility parameter value is selected as the selected node; Performing difference processing on the computing power volatility corresponding to the selected node and the computing power volatility parameter value to obtain computing power volatility difference data between the computing power volatility and the computing power volatility parameter value; Ratio processing is performed on the computing power fluctuation rate difference data corresponding to the selected node and its corresponding computing power fluctuation rate to obtain the computing power fluctuation ratio corresponding to the selected node; Retrieve the task urgency index corresponding to the selected node; wherein the task urgency index is calculated as follows: Task urgency index = number of overdue tasks / total number of processed tasks; Comparing the task urgency index corresponding to the selected node with its corresponding computing power fluctuation ratio; When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the computing power resource threshold is adjusted.

10. The system for building a distributed virtual platform based on blockchain computing power according to claim 9 is characterized in that: When there is a node whose task urgency index exceeds the task urgency index, the computing power resource threshold is adjusted, including: When there is a node whose task urgency index exceeds the computing power fluctuation ratio, the node whose task urgency index exceeds the computing power fluctuation ratio will be used as the target node; The task backlog and utilization gradient values of the target node are retrieved. The task backlog and utilization gradient values corresponding to the target node are obtained using the following formula: When the number of the target node is 1, the task backlog and utilization gradient values corresponding to the target node are used as gradient reference values; When there are multiple target nodes, the standard deviation of the task backlog and utilization gradient values corresponding to the multiple target nodes is used as the gradient reference value; The computing power resource threshold is adjusted using the gradient reference value combined with the spatiotemporal attenuation compensation coefficient corresponding to the target node.

Images