SWIFT: LARGE: Averting Wireless Spectrum Pollution in the Era of Low-Power IoT

点击翻译按钮获取中文摘要

This project seeks to address the problem of spectrum pollution in the Internet-of-Things (IoT) era. Spectrum pollution is an inevitable challenge that emerges when low-cost and low-power wireless IoT devices deployed at scale cannot detect and respect the presence of other devices on shared spectrum. The core challenge is the low-power and simplicity of most IoT devices, due to which they are narrowband and unable to sense and avoid incumbents on shared spectrum. The project investigates a system that allows teams of geo-distributed low-power devices to quickly and efficiently scan wide bandwidths to avert interference. This proposal presents Swallow, a system design for low-power devices to sense spectrum at minimal energy and cost, allowing these devices to behave as low-cost and distributed spectrum observatories. The testbed developed through the project will serve as a vehicle for undergraduate and graduate-level projects as well as workshops for K-12 students in the City of Pittsburgh. The team has direct experience working with sensor deployments both at Carnegie Mellon, the City of Pittsburgh, United States Geological Survey (USGS), and local industry partners and will leverage these connections to deploy Swallow at scale.The project will study a low-power analog frontend and associated digital processing that allows IoT devices to summarize large bandwidths using inexpensive components. It investigates an end-to-end system design that collates measurements from distributed individually low-power IoT devices to obtain a shared spectrum map. It further develops an accurate localization framework to geo-locate individual IoT devices to sample spectrum occupancy at fine-spatial resolution. The wide bandwidth of Swallow’s low-power frontend naturally leads to an accurate solution for time-of flight based localization, where bandwidth is a key metric that dictates system accuracy. Swallow proposes a sub-meter accurate ranging solution for low-power wide-area radios that enables detecting proximity to known passive RF devices such as fixed radio telescopes and enables spectrum occupancy maps sampled at fine-spatial resolution. The project will be implemented and evaluated on a large programmable Low-Power Wide-Area Networking testbed in the Carnegie Mellon University campus that serves large parts of the City of Pittsburgh.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.