Multi-objective Optimization Problem Of Microservice Container Scheduling Based On Improved Particle Swarm Algorithm

In recent years,microservices have been very widely used as a new application development technology in edge computing,Io T,and cloud computing.Application containerization technology is one of its core technologies,which allows multiple containers to be deployed within the same physical node,which makes it possible to run multiple microservices on a single physical node.How to deploy containers rationally on a cluster of physical nodes is one of the main research directions nowadays.Finding a microservice container scheduling solution often requires optimizing multiple objectives,which is a typical multi-objective optimization problem.Although a number of researchers have modeled the microservice container scheduling problem and proposed effective solutions,there are still shortcomings.For example,the model does not accurately describe the multi-objective optimization problem of microservice container scheduling,the search speed is slow,the memory occupation is high,and the algorithm is easy to fall into local optimality.To address the above problems,this paper proposes a Multi-objective Optimization Model based on the Relative Position among Containers(MOM-RPC).Since the particle swarm algorithm has the characteristics of fast convergence,few parameters,and strong optimization finding ability,this paper proposes the Multi-objective Optimization Algorithm based on Parallel Particle Swarm for Microservice Container Scheduling(MOAPPS-MCS)and Particle Swarm Grey Wolf Collaborative Algorithm(PS-GWCA)based on the particle swarm algorithm.The main innovations of this paper are three.(1)MOM-RPC model: Aiming at the modeling of the relationship between containers and their relative positions,which is neglected in the existing models,this model introduces the concept of local load balancing,so that the algorithm can find a better container scheduling scheme when solving the model.Three new optimization objective functions are proposed in three aspects: network transmission overhead,load balancing and service reliability,and these three optimization objective functions are integrated into a new MOM-RPC model.(2)MOAPPS-MCS algorithm: After the particle swarm optimization algorithm is discretized,the representation of individual particles and scheduling scheme is improved for the multi-objective optimization problem of microservice container scheduling,and the update method of particles is improved,so that the particle swarm optimization algorithm is suitable for solving the multi-objective optimization problem of microservice container scheduling.More particle swarm can optimize the solution space through multiple swarm parallel methods,which increases the optimization ability of the algorithm.Experiments show that this algorithm has better and faster optimization efficiency than other algorithms.(3)PS-GWCA algorithm: In this paper,the particle swarm optimization algorithm and the grey wolf algorithm are paralleled.Through the information exchange between the populations,the grey wolf algorithm can guide the particle swarm optimization algorithm to jump out of the local optimization at the early stage of the search,and the particle swarm optimization algorithm at the later stage of the search can enhance the local optimization of the grey wolf algorithm.The experimental results show that the algorithm is significantly better than MOAPPS-CMS algorithm in the optimization results.