Hierarchical and Distributed Software-Defined Network to Reduce Control Load

Abstract

Due to the dissemination of smart devices, such as smartphone, tablet, smartwatch, etc., future mobile network needs to accommodate the mobile traffic volume by 1000-fold [1]. Consequently, the future mobile communication network is required to enhance the capacity. Deployment of smallcell is one approach for capacity increase. The introduction of smallcell brings about the large amount of signaling and management load and the complicated interference management. In order to cope with these two issues, Software-Defined Radio Access Network is proposed [3]. However, Software-Defined Network has a scalability problem. On the other hand, C-plane/U-plane split architecture is proposed as a solution to decrease handover rate and control overhead. However, in C-plane/U-plane split architecture, there is a weak point that the control overhead of control channel of all smallcell user equipments (UEs) is concentrated on macrocell controller and that the controller of macrocell does not scale. In order to address these scalability problems and to enhance capacity, we propose Scalable Software-Defined Network which comprises middlecell. The adaption of middlecell brings about hierarchical and distributed control structure which reduces the control burden of macrocell controller. Moreover, our new algorithm which utilizes both the control channel of macrocell and the control channel of middlecell under in C-plane/U-plane split architecture, keeps handover rate and decreases control overhead. The computer simulation shows the effect of the proposed architecture and the new algorithm about load balance and scalability.