XRShark: Mixed Reality Network Introspection

Abstract

As devices become increasingly mobile and wireless, the last-meter network monitoring problem continues to grow. Network introspection tools will need to support an up-to-date operating picture of dynamic networks, show the physical locations of all network nodes, and connect low-level spatial characteristics with the higher layers of the network stack. Such insights are poorly supported by existing tools that abstract away the physical reality of the underlying network. Fortunately, recent advances in localization technologies and mixed reality platforms provide an opportunity for a new approach: mixed reality network introspection. Mixed reality allows users to see network traffic situated directly in the real world. Physical situation of network activity supports an intuitive understanding of room-scale networks by harnessing human spatial intuition and visual processing, which is well-suited to rapid identification of unexpected activity and correlations in three dimensional space. We propose design goals and a general architecture for mixed reality network introspection, and implement a prototype called XRShark, a network visualizer with augmented reality and virtual reality modes. XRShark enables us to identify several behaviors of the local networks that would be difficult to detect using traditional tools. Based on our prototyping experiences, we discuss promising directions for mixed reality network introspection, as well as remaining challenges for the system design and interaction semantics.

Publication
White Paper
Meghan Clark
Meghan Clark
Postdoctoral Scholar

My research interests include sensor networks, radio communications, and network monitoring.