Towards Distributed SDN: Mobility Management and Flow Scheduling in Software Defined Urban IoT

Di Wu, Xiang Nie, Eskindir Asmare, Dmitri Arkhipov, Zhijing Qin, Renfa Li, Julie McCann, Keqin Li
2018 IEEE Transactions on Parallel and Distributed Systems  
The growth of Internet of Things (IoT) devices with multiple radio interfaces has resulted in a number of urban-scale deployments of IoT multinetworks, where heterogeneous wireless communication solutions coexist (e.g. WiFi, Bluetooth, Cellular). Managing the multinetworks for seamless IoT access and handover, especially in mobile environments, is a key challenge. Software-defined networking (SDN) is emerging as a promising paradigm for quick and easy configuration of network devices, but its
more » ... plication in urban-scale multinetworks requiring heterogeneous and frequent IoT access is not well studied. In this paper we present UbiFlow, the first software-defined IoT system for combined ubiquitous flow control and mobility management in urban heterogeneous networks. UbiFlow adopts multiple controllers to divide urban-scale SDN into different geographic partitions (assigning one controller per partition) and achieve distributed control of IoT flows. A distributed hashing based overlay structure is proposed to maintain network scalability and consistency. Based on this UbiFlow overlay structure, the relevant issues pertaining to mobility management such as scalable control, fault tolerance, and load balancing have been carefully examined and studied. The UbiFlow controller differentiates flow scheduling based on per-device requirements and whole-partition capabilities. Therefore, it can present a network status view and optimized selection of access points in multinetworks to satisfy IoT flow requests, while guaranteeing network performance for each partition. Simulation and realistic testbed experiments confirm that UbiFlow can successfully achieve scalable mobility management and robust flow scheduling in IoT multinetworks; e.g. 67.21% throughput improvement, 72.99% reduced delay, and 69.59% jitter improvements, compared with alternative SDN systems.
doi:10.1109/tpds.2018.2883438 fatcat:dsycq2ymrvgtve2d2jepvvxe6u