Scalable and elastic cloud data center for self-organizing dense small cell networks

Abstract

The exponential growth of wireless data traffic has pushed wireless network vendors and researchers to rethink of a revolutionizing architecture for the fifth generation (5G) wireless network that is capable of supporting 100 times the network capacity of the current 3G/4G systems, yet is able to reduce the net energy consumption of the systems by up to 90 percent. Numerous concepts and technologies have been proposed to approach this goal from various aspects of system design. A consensus reached so far is that 5G is going to be a collection of technology evolutions that include soft core networks, small or even no cellular, pervasive but invisible base stations, and heterogeneous yet self-organizing radio access networks. In addition to advanced radio access technologies, the success of such an ambitious goal really relies on a flexible and scalable network infrastructure to accommodate and harmonize these technologies. On the other hand, virtualization is viewed as one of the essential technologies of the 5G network to provide cost effectiveness of management, maintenance, and hardware upgrade. In this paper, we introduce a cloud service model that has the potential to serve the numerous stringent requirements for future 5G networks. Besides, we describe the designed auto scaling and load balancing rules of each functional block defined in the cloud service model. We conduct emulation to evaluate the performance of resource utilization for the cloud data center.