Construction Of Agent-based Decentralized Grid Model

Grid is an important study in the field of distributed computing.It can combine widely distributed resources and share them all together.It can also eliminate islands of resources to solve large-scale computing problems collaboratively.However,in most grid systems,there is at least one central server to maintain the resource information.As the number of grid nodes increases,the performance bottleneck may be caused.How to obtain resources quickly and accurately in a decentralized grid environment has become a key issue to be solved.Decentralization is a hot research topic in blockchain technology.Its representative PBFT algorithm can reach a consensus through negotiation in a decentralized environment.Applying this algorithm to a grid system helps to share and transfer resources.However,a large number of communication processes in the system may cause efficiency problems.How to improve the system efficiency is an important issue at present.Pi-calculus has powerful ability of formal description and theory of mobile process.It can accurately simulate data transmission and processing in the communication process,especially for parallel operation in distributed systems,and can improve communication efficiency.The n Pict language is a parallel programming language based on Pi-calculus,which can fully implement the operations in the Pi-calculus model and verify its correctness directly.This paper proposes an agent-based decentralized grid model by studying decentralized grid system,mobile agent and theory of Pi-calculus.The PBFT algorithm is used as the research foundation,applied to grid system,and the mobile agent can be used to improve efficiency.Finally,the model is built by Pi-calculus and implemented through nPict programming language to improve efficiency of grid resource sharing in the decentralized environment.The specific work includes the following three points:1.Summarize the study of grid models at present and propose the idea of resource sharing in a decentralized environment.The PBFT algorithm is used as the basic algorithm in this grid model.At the same time,in order to improve the overall performance and execution efficiency of the model,an agent-based decentralized grid model is proposed in combination with the mobile agent technology.2.The agent-based decentralized grid model is built by using Pi-calculus.And to prove the correctness of the model in theory,use the bisimulation theory in Pi-calculus.3.The model is implemented by using nPict language to validate the efficiency and performance problems before and after the mobile agent is added to the model.Through the analysis of the experimental results,it shows that the PBFT algorithm can make the grid service model runs correctly in the decentralized environment,and the efficiency of the model after adding the mobile agent is improved.Therefore,this paper achieves the agent-based decentralized grid model by using Pi-calculus,improving concurrency and execution efficiency of the system,which has important significance for the decentralization of grid resource sharing.