Research On Load Balancing Problem Oriented To Service Quality

The traditional load balancing technology is mainly used to solve the balancing problem of server-side resource allocation,but it not take too much focus on the users' service quality.Therefore,in this respect,it is of great academic significance and application value to study the service quality oriented load balancing problem.On the basis of summarizing the traditional load balancing technology,this paper studies the fair scheduling algorithm based on user service quality and the balanced allocation strategy of resources,the key technology of "load balancing based on service quality",and the theoretical model and key technologies from two aspects: the fair scheduling of service request and the equilibrium assignment of system resources,and finally solves the problem of service quality-oriented load balancing.The paper focuses on the study from three aspects:(1)To make appropriate scheduling strategy on the consider of the characteristics of the request and the resources;(2)To establish a request model which can reflect the characteristics of request and the resource limited;(3)To define the load and the state of equilibrium.The innovation achievements of the research work are reflected in the following aspects:(1)Gives an overall explanation of the load balancing problem.This paper divides the concepts of task scheduling and resource scheduling in load balancing,and makes a deep research on the classification and scheduling algorithm of load balancing technology from the aspects of traditional load balancing strategy,service quality-oriented load balancing strategy and quality of service control.Some qualitative parameters are used to analyze the similarities and differences between the dynamic load balancing algorithm and the static load balancing algorithm,and the commonly used software and hardware load balancing products are briefly analyzed,finally,the problems and challenges in the current load balancing research are summarized.(2)Studies the model for service quality-oriented load balancing.Based on the theoretical model of game theory,this paper presents two models: request scheduling model based on non-cooperative game and the resource allocation model based on cooperative game,from aspects of request scheduling and resource allocation,and then derived the service quality-oriented load balancing model,which provides the theoretical support of the service quality-oriented load balancing problem.(3)Studies the request scheduling based on service quality.To take the improvement of the service quality as the main goal of the request scheduling,this paper proposes a request scheduling strategy based on the satisfaction of service quality through the standardized description of the request model and the service quality,and makes a formal description.In order to meet the demand of service quality,this paper studies the classification method of task request and the queuing theory,and applies the queuing theory model to the request scheduling,and then establishes the request scheduling strategy based on the satisfaction of service qu ality.Finally,the performance and validity of the proposed strategy are analyzed and verified by experiments.(4)Studies the resource allocation strategy based on the maximum entropy principle.Through the study of the optimization theory,the entropy optimization principle and the maximum entropy principle,this paper introduces the concept of entropy into the study of load balancing,and explores the intrinsic relation between entropy and load balance,and proposes the resource allocation strategy based on the maximum entropy principle,which takes the equilibrium allocation of resources as the main goal,and then makes a formal description of its relevant theoretical model.In order to optimize the allocation of load balancing,the multi-attribute load balancing evaluation method and the adaptive load information transmission mechanism are studied,which provides the technical support for the realization of the resource allocation strategy based on the maximum entropy principle.Finally,the concrete description of the strategy is given,and its validity is verified by experimental analysis.In view of the demand for service quality control in load balancing scheduling,this paper presents and studies the model for service quality-oriented load balancing based on the theoretical model of game theory,queuing theory,optimum theory and the maximum entropy principle,and then conducts the research and development of the related key technologies.Through the research work,this paper improves the service quality-oriented load balance theory,and provides relevant technical support,which breaks a new way for the research of load balancing.