This project focuses on the design and implementation of a scalable infrastructure to support a special form of edge computing in which third-party services are deployable in common home gateways in order to support numerous next-generation, low-latency, gigabit applications. Easy availability of such services in the extreme edge of the Internet, enables new capabilities in apps, and services not possible otherwise. The project supports many tangible applications covering different national priority areas (public safety, education, energy, and transportation) across diverse locations (Madison WI, Ammon ID, and Arizona).

The proposed networking infrastructure takes the notion of edge computing to the extreme. It treats the common WiFi Access Point (AP, or a regular home gateway) 'at the ultimate edge of the network' as a true computing platform and provides an edge-cloud abstraction for it. The platform implements unique capabilities through the following three major components: (i) a lightweight virtualization framework on the WiFi AP that allows independent third party services to be deployed and revoked on-demand; implemented with container technologies (LXC, Docker, etc.) we call a container instance in ParaDrop, a chute; (ii) a cloud-based network scaffolding framework that allows third-party developers to deploy and manage such chutes using a Chute Store, in a dynamic and flexible manner and utilizes a software defined networking framework for such management; and (iii) an enhanced API for such chutes that leverages unique knowledge of the wireless context of mobile devices in the vicinity of the APs. The project has ongoing partnerships in multiple locations on applications with direct societal benefits including: (i) creation of a net-zero car-sharing platform in Madison using micro electric vehicles; (ii) a service to provide automated transit analytics to the local bus operation in Madison; (iii) enabling a new mobile service for the Digital Academic Television Network for academic applications in Madison; (iv) a digital public safety service to broadcast robust and real-time safety alerts directly to homes and citizens in Ammon, Idaho; and (v) educational broadcast of augmented reality content to remote classrooms to provide a more immersive learning experience in Arizona.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Type
Standard Grant (Standard)
Application #
1647152
Program Officer
Deepankar Medhi
Project Start
Project End
Budget Start
2017-06-01
Budget End
2021-05-31
Support Year
Fiscal Year
2016
Total Cost
$600,000
Indirect Cost
Name
University of Wisconsin Madison
Department
Type
DUNS #
City
Madison
State
WI
Country
United States
Zip Code
53715