SBIR Phase I:Ultra-Low-Cost Distributed Spectrum Monitoring

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to apply low-cost radiofrequency (RF) sensing hardware to detect, monitor, and localize transmitters in industrial and urban environments. This technology can solve key problems in manufacturing and logistics operations such as finding RF transmitters that are interfering with operations or monitoring the health of radio systems. These activities are required in modern industrial environments as increasing numbers of critical systems rely on radio communications to operate. In urban environments, automating the tasks of detecting, monitoring, and localizing transmitters can simplify the management of large-scale radio networks and gather critical data on wireless device usage. Increased instances of intentional jamming and the rollout of new communications standards such as 5G make these tasks critical to the modern city. Traditionally, such monitoring tasks are conducted manually with expensive spectrum monitoring equipment. Automated installations also utilize expensive sensors, making automated spectrum monitoring only feasible in safety-critical areas such as around airports. Lower-cost sensors can be installed permanently at a high density for almost any application, allowing cities and smaller industrial customers to deploy persistent monitoring networks.This Small Business Innovation Research (SBIR) Phase I project seeks to deploy high-density networks of low-cost sensors, determine the efficacy of existing detection and localization algorithms as applied to the network, and evaluate novel machine-learning (ML) algorithms for similar tasks. This project will also mitigate the technical risk of deploying such a system by characterizing how well commodity software-defined radio (SDR) hardware can perform across a variety of operational environments. No high-density and large-scale test networks of inexpensive radio hardware have been deployed for the purposes of industrial and urban spectrum monitoring, so data gathered throughout this SBIR Phase I project will be useful in evaluating the viability of this approach. By evaluating ML algorithms for signal detection and localization, this project can help determine how effective machine learning models can be at ingesting RF data from low-cost sensors and synthesizing actionable outputs about the radio environment. The result of this project will be an analysis of the detection and localization performance of a variety of algorithms across different environments and sensor node densities.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.