Mixed strategy method for optimizing aggregative indicator under service oriented architecture SOA

Abstract

The present invention discloses an optimizing aggregative indicator mixed strategy method under the service oriented architecture SOA, which belongs to the SOA technical field, and is characterized in that according to demand to each QoS index, the service provider sets the choice probability for the service provider get which is matched and from the service agent, and then a service provider is chosen with the roulette method according to the random number. The method is superior to other methods in the average objective function value, the maximum objective function value, the average waiting time, the maximum waiting time, the average queuing length, and the maximum queuing length.

Description

technical field [0001] The invention belongs to the technical field of SOA. Background technique [0002] Service Oriented Architecture (SOA) is changing the traditional software engineering methods in a loosely coupled, bottom-up, and point-to-point manner to adapt to large-scale, high-complexity market demands; [0003] The service selection problem is a core problem of SOA, which means that the user may find multiple service providers with the same (substitutable) function and different Quality of Service (QoS) through the service broker (Broker). The problem of how to choose the service provider to get the best service; [0004] At present, there are dynamic programming (such as T.Yu, K.-J.Lin.Service selection algorithms for composing complex services with multiple QoS constraints. Lecture Notes in Computer Science, 2005, 3826: 130-143), genetic algorithm (such as H. -C.Wang, C.-S.Lee, T.-H.Ho.Combiningsubjective and objective QoS factors for personalized web service ...

AI Technical Summary

Problems solved by technology

[0006] ·In the peer-to-peer world of SOA, there is no authoritative leader to determine and implement the selection strategy, but the service users determine their own strategies respectively, and all the strategies jointly determine the overall performance of SOA; therefore, the service selection problem It becomes a dynamic game between multiple users with independent decision-making power; however, from the perspective of game theory, the methods in the above literature are all pure strategies, because the equilibrium situation does not necessarily exist for multi-person non-cooperative strategies, The above pure strategy methods often cannot give stable solutions;

[0007] ·Any pure strategy, no matter how good the strategy is, will lead to competition for "high-quality" service providers, thus forming a "winner takes all" situation, that is, most users will choose the service provider that is rated as the best, And other providers will have a greater probability of being idle; if the pure strategy is adopted, this conflict is inevitable; this will not only reduce the QoS index, but also cause uneven load among service providers and a decrease in utilization rate, so that the overall service level drop

Images