| With the diversification of users'requirements, traditional application engines in telecom network have to face more and more challenges, such as the obstruction of business, the high cost of system construction and the complexity of the service provider access. At the same time, Cloud Computing is recognized by the industry gradually due to its high availability, scalability and performance-price ratio. Consequently, the combination of Cloud Computing technologies and the traditional Telecom Network application engine technologies to provide the Application Engine based on Cloud Computing (AECC) has become an important direction. However the introduction of Cloud Computing will bring in a set of uncertainty, i.e. which resource scheduling algorithms could best fit the situation that a great number of applications are deployed into a great number of nodes. The introduction of the Simulation Platform for the Application Engine based on Cloud Computing (SP-AECC) is used to solve the above problems, which is used to simulate and evaluate a variety of configuration options, algorithms and parameters in specific algorithms.Based on the analysis of the evolution of application engines and the mainstream application engines, the application scenarios of AECC are first introduced in this paper. Then the system architecture model of AECC is brought forward. This model abstracts main factors which affect the efficiency of ACEE, such as the model of the node, the resource scheduling algorithm and the task scheduling algorithm, and decouples them from the framework of AECC. By using this model, SP-AECC could change each factor independently. The service model, which is independent of system architecture and supports multiple services and multiple service user models, could be used together with the system architecture model to stimulate different scenarios of AECC. In this paper, the design of SP-AECC is specified. In the design of SP-AECC, the extensibility of the software is emphasized. A set of interfaces are used to facilitate the introduction of new types of nodes, new resource scheduling algorithms and new task scheduling algorithms, which enables the simulation to be implemented more easily and faster and improve the efficiency of the simulation.In the end, the implementation and the tests of SP-AECC are introduced. The front end of SP-AECC is based on javascript, PHP and MySQL, while the back end of SP-AECC is based on OMNet++. And the tests have shown that the platform could not only support correct functionalities, but also performs with good efficiency and stability. |
