Maturing wireless technology and low-cost IoT devices are enabling new kinds of approaches to conventional fields and applications. This project will design a system called IoTScope that aims to shine wireless signals on any object and identify the material properties by analyzing the reflected and refracted signals. Such a capability has the potential to impact many applications helpful to society. Soil moisture detection using drones can help predict crop yield and fertilizer requirement, thus benefitting precision agriculture. Detection of the caloric values of liquids and of water contamination have important health-care applications. Being able to distinguish humans, cars, and other objects can be useful in autonomous-driving and security applications. The project aims to tackle algorithmic and practical challenges in enabling material detection using wireless signals. The simulations, experimental data, and software libraries that will be developed in the project will be made available to students through mini-projects. Given the “Sci-Fi" like nature of this work, it is designed to naturally draw students towards research, while also teaching fundamental principles of wireless communications, signal processing, and wave-matter interaction.
The project attempts to harness the rich information inherent in wireless signals for detecting materials and enabling disruptive IoT applications. Detection can be challenging due to noisy data, wireless multipath (environmental reflections) and diffraction, complexity of material surface, arbitrary user mobility, etc. The specific research tasks are: (I) Fusion of physics models of signal-material interaction with measurements of reflected and refracted signals. This allows IoTScope to decouple signal entanglement with multipath with potentially higher accuracy than state-of-the art multipath deconvolution techniques. (II)Iterative hardware cancellation techniques for subtracting environmental multipath to amplify weak refracted signal. (III) Actuation/mobility based control of probing the material at various angles of incidence, antenna orientations, and frequencies to extract maximal spatial diversity with minimal overhead. (IV) To handle application-specific challenges in soil-moisture detection, the proposal exploits RF-vision fusion together with prior information about soil moisture properties.
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.
