| Due to the high cost of purchasing hardware resources,more and more Internet industries are now investing in the use of IaaS(Infrastructure-as-a-Service)cloud services.Nowadays,the growth of the public cloud market is increasing rapidly.Therefore,research on service pricing strategies and purchasing strategies of IaaS cloud services is important.Its purpose is to provide cloud service providers and users with their own strategies.That is,the provider can make a relatively high profit based on the tasks it receives,and the user can develop a relatively more economical purchasing strategy according to their own usage.Under the development of the cloud market today,this research is significant.In this context,after doing a lot of investigation of the relevant research,we finished some work in the pricing strategy and purchase strategy of the cloud service respectively:1.With respect to service pricing,this thesis develops two kinds of iterative-based approximate optimal pricing algorithms:one is an optimized pricing algorithm based on genetic algorithm,and the other is an optimized pricing algorithm based on hill-climbing method.The former performs better on profitability,but there are higher costs.The latter compensates for the weakness of the former,but it is weaker in terms of profitability than the former.With combining the iterative pricing model,we use meta-heuristic algorithm to solve the complicated multi-variant nonlinear problem to get better solutions.Two kinds of algorithms have shown through a lot of experiments that their profitability is better than the empirical pricing algorithm.2.Regarding the purchase strategy,we use competitive analysis and the new market model,and we develop an online algorithm that can save users' costs more.And we prove that the competitive ratio of the proposed algorithm is lower than the previous literature.Among the different user types,our proposed online purchase algorithm can save more expenses. |
PDF Full Text Request