Edge Computing Based Resource Allocation And Computation Offloading In Wireless Blockchain Network

The decentralization and security characteristics of the blockchain can solve many problems that currently exist in mobile devices and the Internet of Things,such as privacy protection and data security.As the security of the blockchain requires equipment to have huge computing and storage capabilities,many devices currently have insufficient computing and storage capabilities.Therefore,the application of blockchain is severely restricted,and this problem needs to be solved urgently.The proposal of Mobile Edge Computing(MEC)will solve this problem very well.Mobile edge computing sinks from the cloud through user services,which can create a network environment with lower latency,better performance,and greater bandwidth.The combination of MEC and blockchain can not only solve the needs of mobile applications for data security but also solve the problem of insufficient storage and computing power.Therefore,this article mainly studies the resource allocation and computing offloading based on edge computing in the blockchain network.First,this paper designs and implements an edge computing blockchain platform,and then implements two algorithms on this platform.The main tasks of this paper include:Frist:The design and implementation of the edge computing blockchain platform:the introduction of MEC into the blockchain network has brought many new problems.The calculation,storage,and delay characteristics of the blockchain have made the combination of blockchain and MEC new Challenge.Based on the existing research,this paper combines the advantages and disadvantages of the existing research to design and implement a simulation platform.The edge computing blockchain platform is mainly divided into three parts:the back-end part,the algorithm part and the front-end part.The background part uses mainstream background development frameworks for development,and uses docker for application deployment and edge computing-related resource allocation control,and related interfaces are developed to provide complete edge computing blockchain platform functions.In the algorithm part,we deployed related algorithms on the edge computing platform.Finally,we use the front-end framework to build a set of front-end pages to display some data during the operation of the system,so as to visually display the entire system operation process.Finally,we built a monitoring and alarm system for the entire system,which can better monitor the system and ensure the stability of the entire system.Second:Research on edge computing resource allocation strategy in blockchain network:How the MEC server better allocates its own computing and storage resources for blockchain applications is a relatively important issue.The data is transmitted to the edge side,and the relevant blockchain operations are performed on the MEC server.In this way,it can not only ensure the operation of the blockchain network,but also ensure the security and decentralization characteristics of the data.We use the auction algorithm model in economics to treat the resources of edge computing servers as auction items,and the users participating in the competition as bidders.Users first obtain their own estimates based on their task size,computing power,etc.,and algorithms use these estimates to calculate the benefits of edge computing servers.The entire algorithm is completed using a framework based on deep learning algorithms.We built a neural network to simulate the entire auction process and obtain the best results by changing the parameters.The input of the neural network is the bid price of the user,and the output of the neural network is divided into the resource allocation probability and the price that the user pays.Output the probability of a certain user acquiring a certain resource and the price payable respectively.After the optimization of the algorithm,the revenue of the entire system has increased by 20%-30%compared to the benchmark.Compared with the existing single-item algorithm,the profit of the algorithm proposed in this paper has also increased by nearly 25%.Third:Research on computational offloading strategies in blockchain networks:If the tasks of blockchain applications are offloaded to MEC,there will be problems different from traditional offloading methods.First of all,the periodicity of the computing tasks of the blockchain,the high requirements for time delay,etc.,all indicate that the traditional computing offloading method is not suitable for blockchain applications.The difference in blockchain tasks requires a more appropriate calculation offloading algorithm.According to the resource consumption of the blockchain,mining income and so on.We proposed a computational offloading algorithm based on Q learning,constructed a set of action sets,and defined a revenue model.Based on the PID control algorithm can balance the user's income and expenditure and ensure the effective operation of the entire system.In the end,we used two training models,offline training and online training,to optimize the algorithm model.The algorithm can be easily deployed on the edge computing blockchain platform.The experimental results show that in the case of long-term operation,our algorithm can ensure the maximum long-term benefits on the basis of ensuring the stability of the system.After the algorithm is deployed to the edge computing blockchain platform to run,the probability of success of the user's computing blockchain task is increased by more than 60%,which is also 20%higher than other similar studies.