AR over NDN: Augmented reality applications and the rise of information centric networking

Abstract

Collaborative multi-user Augmented Reality (AR) applications pose serious challenges to the underlying network infrastructure due to their all-to-all communication pattern. The Named Data Networking (NDN) paradigm can be a crucial enabler of these applications operated in extremely large scale in terms of users, amount of content, and network size. The inherent multicast support together with a carefully designed naming scheme can provide the efficient network operation, while the inflated Forwarding Information Base (FIB) tables of typical NDN routers can be compressed by powerful algorithms to make the concept feasible. In this demonstration, we showcase an AR application supporting remote collaboration for a large number of users. The software stack of our proof-of-concept prototype leverages open source tools, such as ChronoSync, NDN Forwarding Daemon (NFD), Named Data Link State Routing Protocol (NLSR), and in order to validate the feasibility of the concept, we established a flexible NDN test environment based on Docker containers, real Android clients and emulated users. The framework enables starting arbitrary NDN topologies with predefined FIB contents and to emulate thousands of users. During the live demo, we can show the current network status and relevant performance metrics, such as end-to-end application latency, crucial for AR applications. © 2020 Owner/Author.