Sadp: A State-Aware Domain Partition Approach In Large Scale Software Defined Networking

Jie Gao,Jiangxing Wu,Tao Hu

DOI: https://doi.org/10.1145/3158233.3159352

2015-01-01

Abstract:Distributed multi-domain network architecture has improved the reliability and scalability of software defined networking (SDN). However, it also introduces a new challenge to domain partition because of the uncertainty of network state. In the multi-domain SDN, both excessive and few switch deployments in each domain will bring about the load imbalance of domain controllers and the poor network stability. To address the above problems, for a given large scale SDN network, this paper proposes a State-Aware Domain Partition (SADP) approach. Firstly, through collecting the information of distance, delay and traffic in the network, SADP perceives the network state of all devices, and builds the corresponding state-aware lists of switches and controllers, respectively. Then, according to the specific sequencing of elements in the lists, the awareness connections between switches and controllers are implemented to partition multiple domains and construct the distributed SDN network. Finally, simulation shows that, compared with the typical algorithms, SADP can achieve the better performance of SDN domain partition, which has low processing delay and high load balancing rate.