Edge computing method, system and device based on PaaS cloud platform and medium

Abstract

The invention discloses an edge computing method, system and device based on a PaaS cloud platform and a medium, and the method comprises the steps: carrying out the optimization decision of an edge device through an intelligent decision algorithm, and improving the utilization efficiency of edge computing resources; the security and privacy protection of the edge computing platform are ensured by using a block chain technology and a cryptography technology, and the security of data transmission and storage is ensured; by introducing an automatic operation and maintenance technology, automatic operation and maintenance and management of the edge computing platform are realized, and the management efficiency and stability are improved; by adopting a lightweight virtualization technology, unified management and automatic deployment of edge equipment and a network environment are realized, and the complexity of an edge computing architecture is reduced; the computing power release of edge computing is carried out, and the utilization efficiency and expansibility of edge computing resources are improved through cooperation and resource sharing of a plurality of edge devices. According to the invention, intelligent decision making and optimization on the edge device are realized, and security and privacy protection of the edge computing platform are realized at the same time.

Description

Technical Field

[0001] The present invention relates to the field of edge computing technology, and in particular to an edge computing method, system, device and medium based on a PaaS cloud platform. Background Art

[0002] The edge computing technology of the PaaS cloud platform is a new computing model that has emerged in recent years with the development of technologies such as the Internet of Things and 5G. The traditional cloud computing model mainly concentrates computing and storage resources in cloud data centers, but with the widespread application of the Internet of Things and 5G technologies, more and more data generated by devices and sensors need to be processed and analyzed in real time. For this reason, edge computing technology came into being.

[0003] Edge computing refers to the distribution of computing, storage and other resources to edge nodes close to terminal devices or data sources, so that data can be processed and analyzed locally, thereby achieving low latency, high reliability and high efficiency. The edge computing technology of the PaaS cloud platform is a new technology that applies cloud computing ideas to the field of edge computing.

[0004] The development background of edge computing technology of PaaS cloud platform includes the following aspects: 1) The rapid development of cloud computing, big data, artificial intelligence and other technologies has led to an increase in data volume and computing complexity, requiring more efficient computing and storage resources; 2) The widespread application of Internet of Things technology has required a large number of devices and sensors to perform real-time data processing and analysis, requiring a computing mode with lower latency and higher reliability; 3) The rapid popularization of 5G technology has greatly improved the data transmission speed, making it more suitable for edge computing.

[0005] In practice, the edge computing technology of the PaaS cloud platform needs to solve many problems, including device complexity, network connection quality, data security and privacy protection, cross-device collaboration and resource sharing. In order to solve these problems, the edge computing technology of the PaaS cloud platform adopts many new technologies and methods, such as containerization, multi-path transmission, blockchain, etc. At the same time, the edge computing technology of the PaaS cloud platform also needs to establish unified standards and specifications to facilitate interoperability and collaboration between different manufacturers and devices.

[0006] With the continuous development of technologies such as the Internet of Things and 5G, the edge computing technology of the PaaS cloud platform is expected to be applied in a wider range of fields, bringing more opportunities and challenges to the fields of artificial intelligence, intelligent manufacturing, smart cities, etc. The edge computing technology of the PaaS cloud platform still needs to face many problems and challenges, such as device complexity, network connection quality, data security and privacy protection, cross-device collaboration and resource sharing, etc., and requires continuous technological innovation and exploration. Therefore, the development of the edge computing technology of the PaaS cloud platform still requires the joint efforts of all parties to promote further innovation and application of the technology.

[0007] The defects of edge computing technology in PaaS cloud platform are the heterogeneity of edge devices and network environment, which increases the complexity of edge computing architecture and makes it difficult to meet the needs of complex applications. The security and privacy protection issues of edge computing require better solutions, and the management and maintenance of edge computing platforms require better automation and intelligent support. Summary of the invention

[0008] The purpose of the present invention is to provide an edge computing method, system, device and medium based on a PaaS cloud platform to achieve unified management and automated deployment of edge devices and network environments, realize intelligent decision-making and optimization, improve the utilization efficiency of edge computing resources, and perform security and privacy protection.

[0009] To achieve the above purpose, the technical solution provided by the present invention is:

[0010] The first aspect of the present application provides an edge computing method based on a PaaS cloud platform, including:

[0011] Use intelligent decision-making algorithms to optimize decisions on edge devices and improve the utilization efficiency of edge computing resources;

[0012] Use blockchain technology and cryptography technology to ensure the security and privacy protection of edge computing platforms and ensure the security of data transmission and storage;

[0013] By introducing automated operation and maintenance technology, we can realize automated operation and management of the edge computing platform, and improve management efficiency and stability;

[0014] Use lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture;

[0015] Release the computing power of edge computing, improve the utilization efficiency and scalability of edge computing resources through collaboration and resource sharing of multiple edge devices, and achieve optimal configuration and dynamic adjustment of edge computing resources.

[0016] To optimize the above technical solutions, the specific measures taken also include:

[0017] The mathematical formula for edge computing scenarios where the decision algorithm makes optimized decisions on edge devices is as follows:

[0018] Assume that the task scheduling strategy of the cloud platform is based on the weight of the task and the computing power of the device. Assume that there are N edge devices and M tasks need to be assigned;

[0019] Assume T ij is the execution time of task i on device j, W i is the weight of task i, C j is the computing power of device j, L i is the computational load of task i; the task execution time is expressed by the following formula:

[0020] T ij =L i / C j

[0021] Alternatively, taking into account network latency and inter-device communication overhead, the mathematical formula for the edge computing scenario using an intelligent decision-making algorithm to optimize decisions for edge devices is as follows:

[0022] Assume D ij is the network delay between device i and device j, B ij is the communication bandwidth between device i and device j, S i is the amount of data that needs to be transmitted for task i, L i is the computational load of task i; the task execution time is expressed by the following formula:

[0023] T ij =L i / C j +S i / B ij +D ij .

[0024] The goal of the optimization decision is to optimize the product of the total task execution time and the task weight. The objective function is:

[0025] min∑(W i *T ij )

[0026] The constraints are:

[0027] For each task i, ∑x ij =1

[0028] For task i, whether it is assigned to device j, x ij ={0,1}

[0029] Among them, x ijA binary variable indicating whether task i is assigned to device j.

[0030] Furthermore, the decision tree algorithm and neural network algorithm are used to perform integer linear programming for optimized decision-making, assign tasks and optimize the utilization efficiency and performance of edge computing resources, and use weights to perform weighted combination of the prediction results of the decision tree algorithm and the neural network algorithm; through online learning and adaptive adjustment strategies, deep learning networks and blockchain technology are used to improve the stability and security of the Yunling platform in edge computing.

[0031] Furthermore, in the use of blockchain technology and cryptography technology to ensure the security and privacy protection of the edge computing platform, the SHA-256 algorithm is used as the hash function of the blockchain, the new block is verified in the blockchain through the PoW algorithm, the difficulty target is adjusted according to the computing power of the network, and the integrity of the transaction data in the block is verified through the Merkle tree.

[0032] Block hash calculation:

[0033] Let M be the block data and H be the block hash value. H is calculated by passing M as input to the SHA-256 hash function:

[0034] H=SHA-256(M)

[0035] Proof of Work:

[0036] The PoW algorithm requires the following conditions to be met:

[0037] H<2^(256-n)

[0038] Where n represents the target difficulty; 2^(256-n) represents the maximum range of hash values;

[0039] Difficulty Adjustment:

[0040] Let t be the target block generation time, T be the actual block generation time, n be the current difficulty value, and n' be the adjusted difficulty value. The difficulty adjustment is expressed as:

[0041] n'=n*(t / T)

[0042] Blockchain Verification:

[0043] Each block in the blockchain contains a hash value pointing to the previous block. Let P be the hash value of the previous block, C be the hash value of the current block, H be the hash value of the current block data, and M be the data of the previous block. Then we need to verify the following relationship:

[0044] P = SHA-256 (M)

[0045] C=SHA-256(H)

[0046] In Merkle tree verification, the Merkle root is included in the block header. Let H prev is the hash value of the previous block, M is the Merkle root, N is the random number used for the proof-of-work algorithm, and H block is the hash value of the current block. The block hash is calculated as follows:

[0047] H block =SHA-256(H prev ||M||N).

[0048] The second aspect of the present application provides an edge computing system based on a PaaS cloud platform, including:

[0049] The optimization decision module is used to optimize the edge devices using intelligent decision-making algorithms to improve the utilization efficiency of edge computing resources;

[0050] Security module, which uses blockchain technology and cryptography technology to ensure the security and privacy protection of the edge computing platform and ensure the security of data transmission and storage;

[0051] The operation and maintenance module is used to realize the automated operation and maintenance and management of the edge computing platform by introducing automated operation and maintenance technology, thereby improving management efficiency and stability;

[0052] Virtualization module, which uses lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture;

[0053] The resource configuration module is used to release the computing power of edge computing. Through the collaboration of multiple edge devices and resource sharing, it improves the utilization efficiency and scalability of edge computing resources, and realizes the optimal configuration and dynamic adjustment of edge computing resources.

[0054] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the edge computing method based on the PaaS cloud platform as described in the first aspect of the present application.

[0055] The fourth aspect of the present application provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the edge computing method based on the PaaS cloud platform as described in the first aspect of the present application.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] The present invention applies improved deep learning networks and blockchain technology to the edge computing field of the PaaS cloud platform, realizes intelligent decision-making and optimization on edge devices, and at the same time realizes the security and privacy protection of the edge computing platform; the present invention integrates multiple technologies and applies them to the edge computing field of the PaaS cloud platform, thereby realizing all-round edge computing capability empowerment and improving its stability and security.

[0058] The edge computing technology based on the PaaS cloud platform of the present invention can adopt container technology and lightweight virtualization technology, such as container orchestration technology Kubernetes, to achieve unified management and automated deployment of edge devices and network environments; with the help of decision tree algorithm and neural network algorithm technology, intelligent decision-making and optimization can be achieved on edge devices to improve the utilization efficiency of edge computing resources; blockchain technology and cryptography technology can be adopted, such as data security and privacy protection algorithms of blockchain technology, such as decentralized storage and multi-party computing, to achieve security and privacy protection of edge computing platforms; and the introduction of automated management and intelligent maintenance technology can be adopted to achieve automated operation and management of edge computing platforms.

[0059] The edge computing technology based on the PaaS cloud platform of the present invention can be used in the fields of smart manufacturing, smart cities, smart transportation, the Internet of Things, etc. In these scenarios, devices and sensors need to collect and process a large amount of data, and edge computing can push data processing and analysis to the edge of the network, reduce data transmission and latency, and improve data security and privacy protection; in addition, the edge computing technology of the PaaS cloud platform can also support cross-cloud and multi-cloud architectures to meet the diverse business needs of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 : Flow chart of the edge computing method based on PaaS cloud platform of the present invention.

[0061] Figure 2 : A schematic diagram of the structure of the edge computing system based on the PaaS cloud platform of the present invention. DETAILED DESCRIPTION

[0062] The above contents of the present invention are further described in detail below in the form of embodiments, but this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the above contents of the present invention belong to the scope of the present invention.

[0063] In one embodiment, the present invention proposes an edge computing method based on a PaaS cloud platform. Figure 1 As shown, including:

[0064] Use intelligent decision-making algorithms to optimize decisions on edge devices and improve the utilization efficiency of edge computing resources;

[0065] Use blockchain technology and cryptography technology to ensure the security and privacy protection of edge computing platforms and ensure the security of data transmission and storage;

[0066] By introducing automated operation and maintenance technology, we can realize automated operation and management of the edge computing platform, and improve management efficiency and stability;

[0067] Use lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture;

[0068] Release the computing power of edge computing, improve the utilization efficiency and scalability of edge computing resources through collaboration and resource sharing of multiple edge devices, and achieve optimal configuration and dynamic adjustment of edge computing resources.

[0069] The embodiment of the present invention provides edge computing technology based on the PaaS cloud platform, which mainly includes: intelligent decision-making algorithm, blockchain technology and security, automated operation and maintenance, application of lightweight virtualization technology in edge computing, and realization of edge computing computing power release.

[0070] 1. Intelligent decision-making algorithm

[0071] Decision tree algorithms implement intelligent decision-making and optimization on edge devices, which can effectively improve the utilization efficiency of edge computing resources. Decision tree algorithms predict and make decisions on unknown data by learning and analyzing data. In the field of edge computing, these algorithms can be applied to data analysis, intelligent decision-making, adaptive resource management, etc. to optimize the utilization efficiency and performance of edge computing resources.

[0072] The decision tree algorithm is a classification and prediction algorithm based on a tree structure. Its basic idea is to build a decision tree to predict and make decisions on new data by classifying and analyzing data. In the field of edge computing, the decision tree algorithm can be applied to resource management, load balancing, task scheduling, etc. By analyzing and learning edge devices and network environments, it can realize adaptive allocation and optimal configuration of resources, and improve the utilization efficiency and performance of edge computing resources.

[0073] This application uses the following mathematical formula to describe the resource management and task scheduling problem in edge computing scenarios:

[0074] Assume that the task scheduling strategy of the cloud platform is based on the weight of the task and the computing power of the device.

[0075] Assume there are N edge devices and M tasks need to be assigned.

[0076] Assume T ij is the execution time of task i on device j, W iis the weight of task i, C j is the computing capability of device j.

[0077] In some implementations, the task execution time may be expressed by the following formula:

[0078] T ij =L i / C j ;

[0079] Among them, L i is the computational load of task i.

[0080] The goal is to optimize the product of total task execution time and task weight, that is:

[0081] Objective function: min∑(W i *T ij );

[0082] Constraints:∑x ij =1 (for each task i);

[0083] x ij ={0,1} (indicates whether task i is assigned to device j);

[0084] This is an integer linear programming problem. In practical applications, decision tree algorithm or neural network algorithm can be used:

[0085] (1) Decision Tree Algorithm:

[0086] The decision tree algorithm can be used to predict parameters such as task load, device computing power, network delay, and communication bandwidth. The feature vector X and target variable Y can be constructed based on historical data, where X contains relevant information such as task attributes and device attributes, and Y represents the parameters to be predicted (for example, task load L i , device computing capability C j , network delay D ij , communication bandwidth B ij wait).

[0087] Train a decision tree model:

[0088] Input: feature vector X and target variable Y;

[0089] Output: trained decision tree model;

[0090] Prediction stage:

[0091] Input: Feature vector X of new task and device new ;

[0092] Output: predicted value Y pred .

[0093] (2) Neural network algorithm:

[0094] Neural network algorithms can also be used to predict parameters such as task load, device computing power, network latency, and communication bandwidth. Similar methods can be used to construct feature vectors X and target variables Y.

[0095] Train the neural network model:

[0096] Input: feature vector X and target variable Y;

[0097] Output: trained neural network model;

[0098] Prediction stage:

[0099] Input: Feature vector X of new task and device new ;

[0100] Output: predicted value Y pred .

[0101] (3) Integrate decision tree and neural network algorithms:

[0102] In practical applications, the prediction results of the two algorithms can be weighted and combined to improve the prediction accuracy. Let α be the weight of the decision tree algorithm (0 <= α <= 1), and the weight of the neural network algorithm is 1-α.

[0103] Final predicted value: Y final =α*Y predtree +(1-α)*Y prednn .

[0104] Optimization problem:

[0105] Based on the prediction results, the above optimization problem (integer linear programming problem) can be solved, tasks can be assigned, and the utilization efficiency and performance of edge computing resources can be optimized.

[0106] This mathematical modeling method integrates decision trees and neural network algorithms into the task scheduling problem in edge computing scenarios, and uses these two algorithms to predict and optimize parameters such as task load, device computing power, network latency, and communication bandwidth.

[0107] The algorithm is applied to predict parameters such as task load and device computing power, and then the corresponding optimization algorithm is used to solve the above optimization problem.

[0108] In this example, the meaning of each formula is as follows:

[0109] T ij : The execution time of task i on device j; W i : The weight of task i; C j: computing power of device j; L i : The computational load of task i; x ij : A binary variable indicating whether task i is assigned to device j.

[0110] As a preferred solution, in some implementations, other factors need to be considered, such as network latency, communication overhead between devices, etc., in order to better optimize the utilization efficiency and performance of edge computing resources.

[0111] The previous formula can be modified and expanded to take into account network latency and inter-device communication overhead. New variables are introduced to take these factors into account and the formula is redefined as follows:

[0112] Assume there are N edge devices and M tasks need to be assigned.

[0113] Assume T ij is the execution time of task i on device j, W i is the weight of task i, C j is the computing power of device j, L i is the computational load of task i.

[0114] Assume D ij is the network delay between device i and device j, B ij is the communication bandwidth between device i and device j, S i is the amount of data that needs to be transferred for task i.

[0115] The task execution time can be expressed by the following formula:

[0116] T ij =L i / C j +S i / B ij +D ij ;

[0117] We hope to optimize the product of total task execution time and task weight, that is:

[0118] Objective function: min∑(W i *T ij );

[0119] Constraints:∑x ij =1 (for each task i);

[0120] x ij ={0,1} (indicates whether task i is assigned to device j).

[0121] In this expanded example, each formula means the following:

[0122] Tij : The execution time of task i on device j; W i : The weight of task i; C j : computing power of device j; L i : The computational load of task i; D ij : Network delay between device i and device j; B ij : The communication bandwidth between device i and device j; S i : The amount of data that task i needs to transfer; x ij : A binary variable indicating whether task i is assigned to device j.

[0123] This preferred implementation takes into account network latency and inter-device communication overhead in order to better optimize the utilization efficiency and performance of edge computing resources. In practical applications, decision tree algorithms and neural network algorithms can be applied to predict parameters such as task load, device computing power, network latency, and communication bandwidth, and then the corresponding optimization algorithms can be used to solve the above optimization problems.

[0124] In order to obtain better prediction performance, it is necessary to tune the decision tree and neural network algorithms. For the decision tree algorithm, you can adjust parameters such as the tree depth and node splitting criteria; for the neural network algorithm, you can adjust parameters such as the network structure (number of layers, number of neurons per layer), activation function, and learning rate. By using cross-validation technology, you can evaluate the algorithm performance under different parameter settings to find the optimal parameter combination.

[0125] The prediction results of the decision tree algorithm and the neural network algorithm are weighted and combined using the weight α. In order to determine the appropriate weight, the prediction performance under different weight combinations can be compared experimentally. A method similar to grid search can be used to try different weight combinations within a given range and evaluate their performance using cross-validation techniques. Then, the weight combination with the best performance can be selected.

[0126] In practical applications, parameters such as task load, device computing power, network latency, and communication bandwidth may change over time. In order to make the prediction model adapt to these changes, online learning and adaptive adjustment strategies can be adopted. For example, decision trees and neural network models can be regularly updated and fine-tuned with newly collected data to maintain prediction performance. In addition, the weight α can be dynamically adjusted to assign appropriate weights to prediction results at different time points.

[0127] By tuning decision trees and neural network algorithms, selecting appropriate weights, and online learning and adaptive adjustment strategies, the prediction and optimization effects of task scheduling problems in edge computing scenarios can be further improved, thereby better optimizing the utilization efficiency and performance of edge computing resources.

[0128] By improving deep learning networks and blockchain technology, the Yunling platform can fully empower edge computing capabilities, which increases its stability and security.

[0129] In some embodiments, a deep learning model based on a convolutional neural network (CNN) is designed to predict the task execution time on edge devices. Taking the dataset Google Cluster Data as an example, the dataset provides the task and resource usage of Google data centers. Model training and prediction will be performed based on this dataset.

[0130] Data preprocessing:

[0131] The following features are extracted from Google Cluster Data:

[0132] Task attributes: task type, task priority, CPU and memory required for the task, etc.

[0133] Device attributes: device type, device available CPU and memory, etc.

[0134] Network properties: network latency between devices, communication bandwidth, etc.

[0135] These features are combined into a feature vector X, and the execution time of the task on the device is used as the target variable Y. The dataset is divided into a training set and a test set.

[0136] CNN model design:

[0137] The following is a simple CNN model design:

[0138] Input layer: feature vector X;

[0139] Convolution layer 1: number of convolution kernels: 32, convolution kernel size: 3x3, activation function: ReLU;

[0140] Pooling layer 1: pooling size: 2x2;

[0141] Convolutional layer 2: number of convolution kernels: 64, convolution kernel size: 3x3, activation function: ReLU;

[0142] Pooling layer 2: pooling size: 2x2;

[0143] Fully connected layer 1: Number of neurons: 128, activation function: ReLU;

[0144] Dropout layer: dropout rate: 0.5;

[0145] Output layer: Number of neurons: 1, activation function: linear;

[0146] Model training and evaluation: Use the training set to train the model and the test set to evaluate the model performance. You can try using mean square error (MSE) as the loss function and using stochastic gradient descent (SGD) or Adam as the optimizer. You can optimize the model by adjusting model parameters, number of training iterations, etc.

[0147] Implement intelligent decision-making and optimization on edge devices: The trained CNN model can be deployed on edge devices to predict task execution time. Based on the prediction results, an optimization algorithm (such as integer linear programming) can be used to implement intelligent scheduling and optimization of tasks between edge devices. This can effectively improve the utilization efficiency of edge computing resources.

[0148] Weight quantization is a model compression technology that reduces the storage and computational overhead of the model by reducing the precision of the weights. This method mainly includes the following steps:

[0149] Training model: As before, first train a CNN model;

[0150] Quantize weights: quantize the weights in the model, for example, from 32-bit floating point numbers to 8-bit integers;

[0151] Retraining and fine-tuning: Fine-tune the quantized weights to maintain model performance;

[0152] Deploy quantized model: Deploy the quantized model to edge devices.

[0153] Weight quantization can significantly reduce the storage space and computing overhead of the model while basically maintaining the model performance. Deploying quantized models on edge devices can improve computing efficiency and save resources.

[0154] Network pruning is a technique to reduce model complexity by pruning unimportant neurons or connections to reduce the model's computational overhead. This method mainly includes the following steps:

[0155] Training model: As before, first train a CNN model;

[0156] Evaluate the importance of neurons or connections: The importance of neurons or connections can be evaluated using indicators such as the absolute value of weights and gradients;

[0157] Pruning: Prune a certain proportion of neurons or connections based on the importance evaluation results;

[0158] Retraining and fine-tuning: Retrain and fine-tune the pruned network to maintain model performance;

[0159] Deploy the pruned model: Deploy the pruned model to the edge device.

[0160] Network pruning can effectively reduce the computational overhead of the model while basically maintaining the model performance. Deploying pruned models on edge devices can improve computing efficiency and save resources.

[0161] 2. Blockchain technology and security

[0162] Blockchain technology and cryptography technology, such as blockchain technology's data security and privacy protection algorithms, such as decentralized storage and multi-party computing, realize the security and privacy protection of edge computing platforms.

[0163] Edge computing on the PaaS cloud platform is a new computing architecture that combines cloud computing and edge computing technologies, and uses the collaboration between edge devices and cloud servers to achieve more efficient and reliable computing task processing. In edge computing on the PaaS cloud platform, data security and privacy protection are very important issues, and the use of blockchain technology and cryptography technology is an effective solution.

[0164] Blockchain technology is a decentralized distributed database technology that can realize the sharing and transmission of data between multiple nodes, and has the characteristics of being tamper-proof, decentralized, and distributed. The decentralized characteristics of blockchain technology can avoid single point failures and single point attacks, and improve the security and reliability of data. In addition, the encryption mechanism of blockchain technology can also ensure the security and privacy protection during data transmission.

[0165] Cryptography is a technology that uses cryptographic algorithms to protect data security. In edge computing on the PaaS cloud platform, cryptography can be used to implement data encryption, digital signatures, authentication and other operations to protect data integrity and security. For example, symmetric key or public key encryption algorithms are used to ensure secure data transmission, digital signature algorithms are used to ensure data cannot be tampered with, and identity authentication technology is used to ensure the authenticity of user identities.

[0166] Decentralized storage and multi-party computing are also key technologies applied to edge computing platforms. Decentralized storage can store data in different nodes to avoid single point failures and attacks. Multi-party computing can achieve collaborative computing tasks between multiple edge devices to ensure data security and privacy protection.

[0167] The adoption of these technologies can achieve the security and privacy protection of the edge computing platform of the PaaS cloud platform, ensure the secure transmission and storage of data, and thus improve the reliability and security of the entire computing platform. With the rapid development of technologies such as the Internet of Things and artificial intelligence, edge computing platforms will become increasingly important, and security and privacy protection technologies will also face greater challenges. Therefore, the use of security technologies such as blockchain technology and cryptography technology will become the trend of edge computing development of future PaaS cloud platforms.

[0168] In some embodiments, the blockchain uses the SHA-256 algorithm as its hash function. SHA-256 is a secure hash algorithm that maps data of any length to a fixed-length hash value of 256 bits (32 bytes).

[0169] Block hash calculation:

[0170] Let M be the block data and H be the block hash value. H can be calculated by passing M as input to the SHA-256 hash function:

[0171] H=SHA-256(M)

[0172] Proof of Work (PoW):

[0173] PoW is a consensus algorithm used to verify new blocks in the blockchain. Let H be the block hash value and n be the difficulty target (difficulty value). In the proof of work, H is required to meet the following conditions:

[0174] H<2^(256-n)

[0175] Where n represents the target difficulty, usually an integer. 2^(256-n) represents the maximum range of the hash value.

[0176] Difficulty Adjustment:

[0177] In order to ensure that the block generation rate remains stable, the difficulty target needs to be adjusted according to the computing power of the network. Let t be the target block generation time, T be the actual block generation time, n be the current difficulty value, and n' be the adjusted difficulty value. The difficulty adjustment can be expressed as:

[0178] n'=n*(t / T)

[0179] Blockchain Verification:

[0180] Each block in the blockchain contains a hash value pointing to the previous block. Let P be the hash value of the previous block, C be the hash value of the current block, H be the hash value of the current block data, and M be the data of the previous block, then the following relationship needs to be verified:

[0181] P = SHA-256 (M)

[0182] C=SHA-256(H)

[0183] In blockchain, transaction verification and Merkle tree are two important components. Transaction verification ensures that all transactions are valid, and Merkle tree provides an efficient way to verify the integrity of transaction data in the block. The following will improve the above method based on transaction verification and Merkle tree:

[0184] Transaction Verification:

[0185] In the blockchain, each transaction needs to be verified to ensure the legitimacy of the transaction. Let T be a transaction and the verification function be V(T), then the following conditions must be met:

[0186] V(T) = True;

[0187] Transaction verification usually includes checking for double spending, verifying signatures, etc. Only verified transactions can be added to the block.

[0188] Merkle Tree is a binary tree structure used to efficiently verify the integrity of transaction data in a block. Let ti be the i-th transaction in the block, H(t i ) is t i The hash value of , M is the Merkle root. The construction process of the Merkle tree is as follows:

[0189] a) Calculate the hash value of each transaction: H(t i )=SHA-256(t i );

[0190] b) Use these hash values ​​as leaf nodes of the Merkle tree;

[0191] c) Construct the upper node of the Merkle tree by hashing the adjacent hash value pairs: H(parent) = SHA-256(H(left child )||H(right child ));

[0192] d) Repeat step c until only one node remains, which is the Merkle root: M.

[0193] In the improved block structure, the Merkle root is included in the block header. Let H prev is the hash value of the previous block, M is the Merkle root, N is a random number (used for the proof-of-work algorithm), H block is the hash value of the current block. The new block hash is calculated as follows:

[0194] Hblock =SHA-256(H prev ||M||N).

[0195] The proof-of-work and difficulty adjustment process is the same as before, except that the block hash calculation method has changed. The SHA-256-based blockchain model has been improved by introducing transaction verification and Merkle trees. These improvements make blockchain more suitable for edge computing scenarios, can efficiently verify the integrity of transaction data, and improve the security of the entire system.

[0196] The edge computing of the PaaS cloud platform combines the technologies of cloud computing and edge computing, realizing efficient collaboration between edge devices and cloud servers to improve the processing efficiency of computing tasks. In the edge computing of the PaaS cloud platform, data security and privacy protection are crucial. Blockchain technology and cryptography technology provide an effective way to achieve this goal. Among them, the Merkle tree algorithm is one of the key technologies to achieve blockchain data security.

[0197] The Merkle tree is a binary tree structure used to efficiently verify the integrity of transaction data in the blockchain. It uses the hash value of the transaction data as a leaf node and constructs the upper node by hashing the adjacent hash values. The core advantage of this structure is that only some Merkle tree nodes are needed to verify whether a specific transaction exists in the block. The Merkle tree algorithm uses the SHA-256 hash function to ensure data integrity and security.

[0198] Through the Merkle tree, users can quickly verify whether the transaction data exists in the block without checking the entire block data. This can reduce the computing and storage burden of edge devices. In an edge computing environment, network bandwidth is a precious resource. The Merkle tree allows users to verify transactions by only obtaining partial node data, thereby saving network bandwidth.

[0199] The Merkle tree uses the SHA-256 hash function to ensure data security. It is difficult for attackers to tamper with or forge data in the Merkle tree, thereby improving the security of the entire system. The Merkle tree structure can speed up transaction confirmation and improve the performance of edge computing systems.

[0200] The Merkle tree algorithm has important application value in the edge computing scenario of the PaaS cloud platform. It can efficiently verify the integrity of transaction data, save network bandwidth, improve system security, and accelerate transaction confirmation. By applying blockchain technology, cryptography technology, and the Merkle tree algorithm to the edge computing of the PaaS cloud platform, data security and privacy protection can be ensured, and the reliability and security of the entire computing platform can be improved.

[0201] 3. Automated operation and maintenance

[0202] Introduce automated management and intelligent maintenance technologies to realize automated operation and management of the edge computing platform, and improve the management efficiency and stability of the edge computing platform.

[0203] The introduction of automated management and intelligent maintenance technologies is an important means to solve the management and maintenance problems in edge computing of PaaS cloud platforms. With the increase in the number of edge devices and the complexity of application scenarios, manual management and maintenance of edge computing platforms have become increasingly difficult. Automated management and intelligent maintenance technologies can achieve automated operation and management of edge computing platforms through automated tools, machine learning and other technical means, thereby improving management efficiency and stability.

[0204] Automated management technologies include automated deployment, automated monitoring, and automated scaling, among which automated deployment is the most basic technical means. Automated deployment can be achieved through orchestration tools (such as Kubernetes) and automated scripts, thereby enabling rapid deployment and configuration of edge computing platforms. Automated monitoring can be achieved through monitoring indicators and alarm mechanisms to promptly discover and resolve problems in edge computing platforms. Automated scaling can automatically adjust computing resources according to actual load conditions, thereby improving resource utilization.

[0205] Intelligent maintenance technologies include automated troubleshooting and automated optimization. Automated troubleshooting can quickly locate the cause of a fault through machine learning and other technical means, thereby improving the efficiency of troubleshooting. Automated optimization can improve the performance and efficiency of the edge computing platform through algorithm optimization, resource adjustment and other technical means, thereby improving the stability and reliability of the entire platform.

[0206] The introduction of automated management and intelligent maintenance technologies is crucial to the stability and reliability of the edge computing platform of the PaaS cloud platform. These technical means can realize the automated operation and management of the platform, reduce the time and workload of manual intervention, and improve management efficiency and stability.

[0207] 4. Application of lightweight virtualization technology in edge computing

[0208] Adopt container technology and lightweight virtualization technology, such as Kubernetes and Docker, to achieve unified management and automated deployment of edge devices and network environments, and reduce the complexity of edge computing architecture.

[0209] The edge computing technology of the PaaS cloud platform uses container technology and lightweight virtualization technology, such as Kubernetes and Docker, to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture. Container technology is a lightweight virtualization technology that uses the characteristics of the operating system to encapsulate the application and all its dependencies in a container to achieve rapid deployment and cross-platform porting. Kubernetes is an open source container orchestration system that provides a scalable, highly available containerized application management platform to help users automate the deployment, management, and expansion of containerized applications. Docker is an open source containerization platform that provides a portable, scalable containerized environment to help users quickly build, test, and deploy applications.

[0210] The application of container technology and lightweight virtualization technology can reduce the complexity of edge computing platforms and improve their reliability, scalability and flexibility. These technologies can help users manage edge devices and network environments more easily, quickly deploy and upgrade applications, and improve development and deployment efficiency. At the same time, they can also improve the reliability and stability of edge computing platforms and ensure their high availability and fault tolerance.

[0211] Container technology and lightweight virtualization technology play a vital role in edge computing on PaaS cloud platforms. Their application can improve the reliability, scalability and flexibility of edge computing platforms, reduce the complexity of edge computing platforms, and help users better cope with the growing demand for edge computing.

[0212] 5. Unleash the computing power of edge computing

[0213] Improve the utilization efficiency and scalability of edge computing resources, and achieve optimal configuration and dynamic adjustment of edge computing resources through multi-device collaboration and resource sharing.

[0214] Improving the utilization efficiency and scalability of edge computing resources is an important task in the edge computing of PaaS cloud platforms. In the edge computing environment, devices and resources are widely distributed and highly heterogeneous. How to better coordinate and utilize these distributed resources to improve computing efficiency and scalability is one of the difficult problems that the edge computing technology of PaaS cloud platforms needs to solve.

[0215] To solve this problem, we can achieve optimal configuration and dynamic adjustment of edge computing resources through multi-device collaboration and resource sharing, etc. The technologies involved include distributed computing, virtualization technology, container technology, load balancing, etc.

[0216] Distributed computing refers to breaking down a large computing task into multiple subtasks, then assigning these subtasks to different computing nodes for parallel computing, and finally merging the results. In an edge computing environment, distributed computing technology can assign computing tasks to multiple edge devices for parallel computing, improving computing efficiency and performance.

[0217] Virtualization technology and container technology can achieve virtualized management and isolation of edge device resources, improving the utilization of edge computing resources. Through virtualization technology, physical resources on edge devices can be abstracted into virtual resources, thereby achieving unified management of computing, storage, network and other resources. Container technology can achieve resource isolation and management in a more lightweight manner, providing a more flexible deployment and management method for edge computing.

[0218] Load balancing technology is also one of the key technologies to improve the utilization and scalability of edge computing resources. Through load balancing technology, the computing load on edge devices can be distributed to different computing nodes for processing, thereby avoiding excessive load on a certain node, which may lead to system performance degradation or crash. Common load balancing algorithms include polling algorithm, minimum connection number algorithm, minimum response time algorithm, etc.

[0219] Improving the utilization efficiency and scalability of edge computing resources requires the use of a variety of technical means, including distributed computing, virtualization technology, container technology, load balancing technology, etc. These technical means can make better coordinated use of edge computing resources, improve computing efficiency and performance, and thus achieve wider and deeper application of edge computing application scenarios.

[0220] The above-mentioned intelligent deep learning technology solves the problem of edge computing resource optimization, and encryption solves the platform security problem. These technical solutions provide strong technical support for the Yunlingyun platform.

[0221] The main feature of the edge computing technology of the PaaS cloud platform of the Yunling container cloud platform in this application is the use of container technology and lightweight virtualization technology, such as Kubernetes and Docker, to achieve unified management and automated deployment of edge devices and network environments.

[0222] Compared with traditional methods, Yunling Container Cloud Platform uses container technology to provide more efficient and flexible edge computing services, while achieving unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture.

[0223] Yunling Container Cloud Platform also uses machine learning and data analysis technologies to implement intelligent decision-making and optimization on edge devices, improving the utilization efficiency of edge computing resources; it uses blockchain technology and cryptography technology, such as blockchain technology's data security and privacy protection algorithms, such as decentralized storage and multi-party computing, to achieve security and privacy protection of the edge computing platform.

[0224] This application introduces automated management and intelligent maintenance technologies to realize automated operation and management of the edge computing platform, making the edge computing technology of the PaaS cloud platform of Yunling Container Cloud Platform more efficient, secure, and intelligent, helping enterprises achieve better business efficiency and security in digital transformation.

[0225] In another embodiment, the present invention proposes an edge computing system based on a PaaS cloud platform, such as Figure 2 As shown, including:

[0226] The optimization decision module is used to optimize the edge devices using intelligent decision-making algorithms to improve the utilization efficiency of edge computing resources;

[0227] Security module, which uses blockchain technology and cryptography technology to ensure the security and privacy protection of the edge computing platform and ensure the security of data transmission and storage;

[0228] The operation and maintenance module is used to realize the automated operation and maintenance and management of the edge computing platform by introducing automated operation and maintenance technology, thereby improving management efficiency and stability;

[0229] Virtualization module, which uses lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture;

[0230] The resource configuration module is used to release the computing power of edge computing. Through the collaboration of multiple edge devices and resource sharing, it improves the utilization efficiency and scalability of edge computing resources, and realizes the optimal configuration and dynamic adjustment of edge computing resources.

[0231] In another embodiment, the present invention proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, an edge computing method based on a PaaS cloud platform as described in the present application is implemented.

[0232] In another embodiment, the present invention provides a computer-readable storage medium storing a computer program, which enables a computer to execute an edge computing method based on a PaaS cloud platform as described in the present application.

[0233] In the embodiments disclosed in the present application, the computer storage medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. The computer storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the above. More specific examples of computer storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0234] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in any form. Any technician familiar with the profession, without departing from the scope of the technical solution of the present invention and based on the technical essence of the present invention, may make any simple modifications, equivalent substitutions and improvements to the above embodiments, which are still within the protection scope of the technical solution of the present invention.

Claims

1. An edge computing method based on PaaS cloud platform, characterized in that: include: Use intelligent decision-making algorithms to optimize decisions on edge devices and improve the utilization efficiency of edge computing resources; Use blockchain technology and cryptography technology to ensure the security and privacy protection of edge computing platforms and ensure the security of data transmission and storage; By introducing automated operation and maintenance technology, we can realize automated operation and management of the edge computing platform, and improve management efficiency and stability; Use lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture; Release the computing power of edge computing, improve the utilization efficiency and scalability of edge computing resources through collaboration and resource sharing of multiple edge devices, and achieve optimal configuration and dynamic adjustment of edge computing resources.

2. The edge computing method based on the PaaS cloud platform according to claim 1 is characterized in that: The mathematical formula for the edge computing scenario of using intelligent decision-making algorithms to optimize decisions for edge devices is as follows: Assume that the task scheduling strategy of the cloud platform is based on the weight of the task and the computing power of the device. Assume that there are N edge devices and M tasks need to be assigned; Assume T ij is the execution time of task i on device j, W i is the weight of task i, C j is the computing power of device j, L i is the computational load of task i; the task execution time is expressed by the following formula: T ij =L i / C j Alternatively, taking into account network latency and inter-device communication overhead, the mathematical formula for the edge computing scenario using an intelligent decision-making algorithm to optimize decisions for edge devices is as follows: Assume D ij is the network delay between device i and device j, B ij is the communication bandwidth between device i and device j, S i is the amount of data that needs to be transmitted for task i, L i is the computational load of task i; the task execution time is expressed by the following formula: T ij =L i / C j +S i / B ij +D ij 。 3. The edge computing method based on the PaaS cloud platform according to claim 2 is characterized in that: The goal of the optimization decision is to optimize the product of the total task execution time and the task weight. The objective function is: min∑(W i *T ij ) The constraints are: For each task i, ∑x ij =1 For task i, whether it is assigned to device j, x ij ={0,1} Among them, x ij A binary variable indicating whether task i is assigned to device j.

4. The edge computing method based on the PaaS cloud platform according to claim 1, characterized in that: The decision tree algorithm and neural network algorithm are used to optimize integer linear programming for decision making, assign tasks and optimize the utilization efficiency and performance of edge computing resources, and use weights to perform weighted combination of the prediction results of the decision tree algorithm and the neural network algorithm; through online learning and adaptive adjustment strategies, deep learning networks and blockchain technology are used to improve the stability and security of the Yunling platform in edge computing.

5. The edge computing method based on the PaaS cloud platform according to claim 1, characterized in that: In the use of blockchain technology and cryptography technology to ensure the security and privacy protection of the edge computing platform, the SHA-256 algorithm is used as the hash function of the blockchain, the new block is verified in the blockchain through the PoW algorithm, the difficulty target is adjusted according to the computing power of the network, and the integrity of the transaction data in the block is verified through the Merkle tree.

6. The edge computing method based on the PaaS cloud platform according to claim 5 is characterized in that: Block hash calculation: Let M be the block data and H be the block hash value. H is calculated by passing M as input to the SHA-256 hash function: H=SHA-256(M) Proof of Work: The PoW algorithm requires the following conditions to be met: H<2^(256-n) Where n represents the target difficulty; 2^(256-n) represents the maximum range of hash values; Difficulty Adjustment: Let t be the target block generation time, T be the actual block generation time, n be the current difficulty value, and n' be the adjusted difficulty value. The difficulty adjustment is expressed as: n'=n*(t / T) Blockchain Verification: Each block in the blockchain contains a hash value pointing to the previous block. Let P be the hash value of the previous block, C be the hash value of the current block, H be the hash value of the current block data, and M be the data of the previous block. Then we need to verify the following relationship: P = SHA-256 (M) C=SHA-256(H) In Merkle tree verification, the Merkle root is included in the block header. Let H prev is the hash value of the previous block, M is the Merkle root, N is the random number used for the proof-of-work algorithm, and H block is the hash value of the current block. The block hash is calculated as follows: H block =SHA-256(H prev ||M||N)。 7. An edge computing system based on a PaaS cloud platform, characterized in that: include: The optimization decision module is used to optimize the edge devices using intelligent decision-making algorithms to improve the utilization efficiency of edge computing resources; Security module, which uses blockchain technology and cryptography technology to ensure the security and privacy protection of the edge computing platform and ensure the security of data transmission and storage; The operation and maintenance module is used to realize the automated operation and maintenance and management of the edge computing platform by introducing automated operation and maintenance technology, thereby improving management efficiency and stability; Virtualization module, which uses lightweight virtualization technology to achieve unified management and automated deployment of edge devices and network environments, reducing the complexity of edge computing architecture; The resource configuration module is used to release the computing power of edge computing. Through the collaboration of multiple edge devices and resource sharing, it improves the utilization efficiency and scalability of edge computing resources, and realizes the optimal configuration and dynamic adjustment of edge computing resources.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the edge computing method based on the PaaS cloud platform as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the edge computing method based on the PaaS cloud platform as described in any one of claims 1 to 6.

Images