A Stepwise Self-adaptive Model for Improving Cloud Efficiency Based on Multi-Agent Features

Today, the multi-agent systems are the most common systems that have intelligent behavior and able to adapt features based on the environmental changes. On the other side, cloud applications enable the stakeholders to customize their resources and software they need based on the requested domain. These applications face many challenges such as how to handle the changes of the stakeholder requirements at run-time, how to reconfigure the constituted architecture dynamically to be in consistency with the new services, and how to cope with the highly inherent expensive cost. To deal with these challenges, we proposed a new model that uses basic agent features for enhancing the cloud infrastructure functionalities by reconfiguring their allocated resources and software at run-time. The proposed model is composed of three levels. The first level is the cloud level which via its functionalities the consistent image that maps the user's requests through its manual components can be created & established. The second level is the intermediate level, which is responsible for two issues: playing the role of connector between the cloud level and the multi-agent level, and verifying the consistent of outputs for both of the upper and lower levels. The third level is the multi-agent, which is responsible for improving the quality of the constituted cloud images by co-operating the information, reasoning, learning and mobile agents. Finally, the urban transportation system is used to proof the applicability and usage of the proposed model.