A New Architecture of an Autonomous System in Cloud Computing

The continuous increase of network devices complicates the management task of appliances and services. The complexity and the dynamism of the current global network system require an interacting architecture. This should be able to autonomously change the system’s structure and functionality with little human intervention. Moreover, the cloud elastic nature necessitates the frequent need to a dynamic reconfiguration of the offered services. This is useful in case of unpredictable failure such as system crashes, network problems, or natural disasters that typically cause service unavailability. This paper proposes a new network system architecture based on cloud computing for service management. The suggested architecture is built using two architectures; namely, the autonomic SOA framework and the distributed autonomous control framework. The autonomic SOA framework extends the intelligence and capability in the cloud by using case-based reasoning as well as the architectural consideration of autonomic computing paradigm. On the other hand, the main feature of the second framework is the classification of the services into basic real-time and sophisticated non-real-time applications. The former are realized through M2M (Machine-to-Machine) communication between the devices under the local network area; whereas the Internet manages the non-real-time applications and controls the supporting devices.