Net SE: Medium: Collaborative Research; Towards Human-Network Interaction (HNI) for the Home

As networks move outside professionally-managed environments, end-users-who may have neither inclination towards nor formal training in network design and management-become nonetheless tasked with architecting, deploying, managing, securing, and troubleshooting their networks. This shift has beenmarked by increasing evidence that end-users cannot manage networks, and their rejection of these technologies represents a significant barrier to the adoption of advanced applications. This proposed research aims to lower the barrier, arguing that the challenges of human-network interaction in end-user\managed environments presents an important opportunity to explore new technical approaches that solve-or at the very least mitigate-these problems. The Intellectual Merit of this research lies in its inherently cross-disciplinary focus, bringing HCI research to bear on and informing the science and engineering of networking in end-user managed environments. This approach is inherently socio-technical, examining the science of networking not solely from a technical perspective, but from one with an understanding of the lived experience of networking in end-users' hands. The key contributions that we aim for in this work include: Providing a much deeper understanding of the experience of networking in end-user managed environments. Through ethnographic and other empirical methods, we will produce a dataset and associated analyses that detail the challenges of networking in such environments, the intentions and needs of the users who both manage and live with such networks, and the resulting opportunities for design that can inform the creation of future network architectures and protocols. Creating a novel testbed architecture for exploration of end-user managed networks. The initial version of this architecture, derived from principles developed as part of our earlier work, centralizes switching, policy enforcement, and control logic on the network, and isolates users from many of the tedious (and error-prone) aspects of provisioning and configuration. We will extend this architecture, deploy it in real users' homes, and evaluate it from both technical and user-centric perspectives. Producing sound, generally-applicable principles for more usable network design. These principles will be particularly applicable to contexts where no formal administrator is present. While our central focus is on the home, we believe that many of our results-such as guidelines for creating more expressive policy primitives, or visual tools for network management-will be generalizable and more broadly applicable outside of the home setting. This project, aims to break ground in a new area that might be called Human-Network Interaction: the science of understanding how to create network architectures and tools that are more approachable and understandable by their users, and a better fit for their needs and intentions. The intellectual challenge of this area comes, first, from the collection and analysis of empirical data in order to gain insight into the structure of problem and solution spaces, and their relationship to each other; and second, from the invention, evaluation and refinement of solutions in the relevant parts of the space. Human-network interaction, it is suggested, is one component of Networked Science and Engineering (NetSE), the science and engineering of networking technologies that empower and engage end-users. This proposal has the following Broader Impacts. On Society: Usable home networking is on the critical path for computing research innovation. On Education: The results of our proposed project will teach the next generation of engineers how to design more usable networking technology. On Underrepresented Groups: The integration of usability concerns into the core of computing reflects the growing diversity of careers in IT, and opens up the possibility of increasing the participation of women and minorities in computing. On Research: