Routing in Large-scale Dynamic Networks

Abstract

The increasing volume of network-connected devices comprising Internet of Things and the variety of heterogeneous network architectures across these devices pose significant challenges to effective deployment and routing. In this article, we consider the adoption of probabilistic data structures to develop a novel Bloom Filter-based dual-layer inter-domain routing scheme. Our designed scheme implements internal and external routing layers in network gateways constructed upon the counting bloom filter and the original bloom filter. We first compare several representative structures in both theory and experimentation. We then propose our novel Bloom Filter-based dual-layer inter-domain routing scheme. In the design of the routing scheme, we consider issues related to the overall space cost and routing loop prevention, as well as present corresponding solutions. We also detail the principal structures and algorithms. Further, we conduct a theoretical analysis of the space efficiency of our proposed scheme compared to traditional routing with respect to the size of data packets and the size of routing tables, as well as in routing loop avoidance. Finally, via extensive performance evaluation, our experimental results demonstrate the effectiveness and efficiency of our proposed scheme.