Supporting pub/sub over NDN sync

Abstract

Distributed dataset synchronization, or Sync, plays the role of a transport layer protocol in Named Data Networking (NDN). The role of Sync is to synchronize the namespace of all data productions by multiple entities running the same application. NDN application developers desire a high level API, such as the commonly used pub/sub API that hides transport and network layer details. This poster explores the design of such an API built on top of State Vector Sync (SVS), one of the NDN Sync protocols, along with a low-latency data fetching option. With this API, SVS provides fast and resilient dataset synchronization, enabling developers to work with a familiar pub/sub API while benefiting from NDN's capabilities of built-in data security, multicast data delivery, in-network caching, and consumer driven flow control.