Performance improvement for source mobility in named data networking based on global–local FIB updates

Abstract

The mobile devices have become the norm, rather than an exception for people’s daily life, and these devices possess strong ability (e.g., increasingly large storage) to serve as content sources in future information-centric networks, e.g., named data networking (NDN). The efficient discovery of contents provided by mobile devices is a challenging task. Most of existing studies adopted ID-Locator split architecture to handle the problems raised by source mobility in NDN, e.g., the content source is suddenly disconnected from the network when providing on-going services. However, these solutions require very high control overhead to maintain up-to-date location information of sources in an ID-Locator mapping system. Software defined networking (SDN) is a promising tool to make cross-layer decisions based on its logically centralized controller, and thus can facilitate the implementation of forwarding strategies in NDN according to application requirements. To this end, this paper proposes a new packet forwarding algorithm under SDN paradigm that is able to update the forwarding information base (FIB) of NDN node considering both the mobility pattern of mobile sources and network states. In particular, the algorithm leverages the SDN controller to perform the tradeoff between global FIB updates on all NDN nodes and local FIB updates on NDN nodes that are previously connected by mobile sources to lower the control overhead. A discrete-event simulator based on NS3 simulation framework is developed to validate the effectiveness and feasibility of the proposed algorithm. In addition, the performance of the algorithm is compared with that of existing solutions based on ID-Locator split architecture. The results show that the proposed packet forwarding algorithm has a lower request-to-response latency and higher successful content request ratio with reduced control overhead under different working conditions.