Optimizing Bi-directional Low-Latency Communication in Named Data Networking

TL;DR

This paper proposes architectural enhancements to Named Data Networking (NDN) to improve throughput and reduce latency for low-latency, bidirectional communication applications, validated through simulations and testbed experiments.

Contribution

It introduces specific architectural amendments to NDN tailored for low-latency, bidirectional communication, addressing a gap in NDN's suitability for non-content-centric applications.

Findings

Significant improvement in throughput for bidirectional applications.

Reduced latency achieved through proposed architectural modifications.

Validated results from both simulations and testbed experiments.

Abstract

Content-Centric Networking (CCN) is a concept being considered as a potential future alternative to, or replacement for, today's Internet IP-style packet-switched host-centric networking. One factor making CCN attractive is its focus on content distribution, which dominates current Internet traffic and which is arguably not well-served by IP. Named Data Networking (NDN) is a prominent example of CCN. It is also one of several on-going research efforts aiming to design and develop a full-blown candidate future Internet architecture. Although NDN's primary motivation is content distribution, it is envisioned to support other types of traffic, such as conferencing (audio, video) as well as more historical applications, such as remote login. However, it is unclear how suitable NDN is for applications that are not obviously content-centric. In this paper, we explore NDN in the context of a class of applications that involve low- latency bidirectional communication. Specifically, we propose a few architectural amendments to NDN that provide significantly better throughput and lower latency for this class of applications. The proposed approach is validated via both simulations and testbed experiments.