Principal Investigator: Professor Fay

Other Contributor: Nivedhita Venkatesan

AWaRE REU Researcher: Fuwei Huang, University of Colorado Denver

Project Description: Novel transistor designs for improved linearity in GaN-based FETs are being explored for their potential to improve dynamic range in mm-wave low noise amplifiers (LNAs). This project includes device design, modeling, fabrication, and characterization of devices, as well as the design of low-noise amplifiers (based on the extracted models) and comparison with designs based on conventional transistors in order to fully understand the potential benefits and any associated design trade-offs for mm-wave receiver applications.

Finding: In the world of wireless communication, HEMT’s (High Electron Mobility Transistors) play a critical role in the recent development. Because of the unique property of high power, high frequency, and high gain, it makes a great candidate for military and commercial wireless applications.

For cellar station applications, GaN transistor device is more efficient at higher frequency with broader bandwidth than the Si LDMOS (Silicon laterally diffused metal oxide semiconductor device). Which in terms, give us a higher data transfer rate. Because of the higher bandgap of Gallium Nitride semiconductor material, GaN HEMT is the perfect replacement for the GaAs pHEMTs as it could achieve higher power, higher frequency, and higher bandwidth at the same time.

The goal of our research is to develop a GaN device model as an amplifier that will be able to satisfy our ever-advancing technology needs in the next few decades. As of now, we are in the early development phase where we model our device in the Synopsys Technology Computer-Aided Design (TCAD ) and optimize it in the Advanced Design System (ADS) for higher gain, OIP3, and IIP3 value. The research included doing analytical Figures of Merit for the model device and experimenting with Low Pull simulation as an alternative, more efficient way to get load impedance from the model, which speed up the research progress.

Principal Investigator: Professor Hochwald

Other Contributor: Arash Ebadi Shahrivar

AWaRE REU Researcher: Zachary Schoon, Bethel University

Project Description: RadioHound, an ongoing project at NDWI, is the development of low-cost, portable spectrum measurement sensors capable of tuning over a wide range of frequencies commonly used by everything from cellular phones to wireless local area networks, to radios and televisions. One goal is to distribute these sensors over a wide geographical area and thereby crowd-source the real-time measurements to create a “heat-map” of spectrum usage over the area and across frequency. Such a map would be used, for example, to determine where spectrum congestion is dense.

We are in the fourth year of this project.

Finding: Spectrum sensing is the process of periodically monitoring a specific frequency band, aiming to identify the presence or absence of primary users. The RadioHound spectrum sensor is able to tune into a wide range of frequencies that are used by radios, WiFi, television, and cellular devices. For the sensors to output information pertaining to network activity, wireless network packets must be decoded.

Our team has created a MATLAB code that will interpret WiFi network packets, decode them, and provide the output in a PCAP file. The issue with our written code is that it cannot run natively on the RadioHound device because of its dependence on Matlab.

Our goal is to be able to tune to a 20 MHz bandwidth signal with the RadioHound device and demodulate any WiFi signal contained therein. To accomplish this, we have translated the Matlab code to Python, which can run natively on the RadioHound device. Once the Python code is completed, we will be able to identify and demodulate WiFi data readily.

Principal Investigator: Professor Lin

Other Contributor: Wei Zheng

AWaRE REU Researcher: Chase Brown, Bethel University

Project Description: In Human-Robot Collaboration, robots are expected to work next to humans in warehouses, daily housekeeping, and other robot assistant applications safely, intelligently and friendly. To achieve this goal, the robotic system should be equipped with capacities of understanding the intentions of human partners and reasoning according to the behaviors of human partners and the state of the environment. The main idea is to combine the learning-based approach with traditional high-level task planning algorithms. The first step is to build a human model using data collected from the visual perception system such as stereo cameras. Based on the learned human model, robots could infer the intentions of human partners using the data collected during run-time. For example, in the handover task, the robot could track the skeleton of the human, collect data from several demonstrations and then infer the human intention. Once the robots understand human intention, they could behave collaboratively with the human according to decisions made by high-level task planning algorithms.

Finding: The collaboration of humans and robots in the industrial and home environment is essential to the progression of modern robotics. The implementation of a vector autoregressive partially observable Markov decision process (VAR-POMDP) allows humans and robots to collaborate on high-level objectives with precision and accuracy.

The undergraduate research experience in human-robot collaboration is tasked with the implementation of this model in a low-cost personal robot. Creating a system for tracking a unique stochastic (human) model is integral for providing meaningful data to the robotic control system.

The VAR-POMDP learning model is recognized for its ability to accurately predict the stochastic actions of a human contained within a predefined environment. Distinct from prior work the proposed demonstration does not predefine human states or transition states in order to show the flexibility of the proposed VAR-POMDP model in many unique situations.

A Bayesian non-parametric learning model is implemented in the construction of the demonstration to define potential human states through observable data collected within the operating environment. The states defined by the Bayesian learning model are multiplied by the predefined discretized states of the robotic environment to create a set of all perceivable states. These states form the foundations for a state transition matrix which is used to predict human behavior. The successful implementation of this demonstration has advanced the validity of the proposed VAR-POMPD model. The further implementation of this model in the fields of assisted living and driver assistance systems will induce innovation and advancements.

Principal Investigator: Professor Chisum

Other Contributor: David Connelly

AWaRE REU Researcher: Muhammad Hussain, Stony Brook University

Project Description: Reconfigurable antennas and distributed circuits have become extremely relevant for today’s extremely wideband spectrum operations. Military applications require spectrum sensing from HF to 100’s of GHz, and commercial industry is interested in leveraging software-defined radios covering the “DC” to 6 GHz band. In either case, recent advances in transistors can provide low-pass frequency response covering DC to 100’s of GHz. However, at some point these electronics must interface with antennas whose dimensions are fundamentally tied to the frequency of operation, they are band-pass structures. For this reason, single antennas cannot provide wideband frequency coverage to match that of the electronics.

We have recently demonstrated 1D programmable transmission lines based upon metallic inclusions in a vanadium dioxide (VO2) film with low on-state loss and high off-state isolation. Such a material enables programmable antennas that are capable of matching the bandwidth of the electronics and can, therefore, offer revolutionary solutions to wireless sensing and communications. This project will extend the 1D proof-of-concept to a 2D programmable material for the purpose of antenna applications.

Finding: Reconfigurable antennas have become increasingly relevant for wideband spectrum operations as the frequency of distributed circuits corresponds to their physical geometry. Currently, advanced RF ICs enable extremely wideband radios, but lack corresponding antennas and other distributed structures. Reconfigurable distributed circuits have been realized with switch networks (e.g., antenna array, switch matrix) which limits flexibility.

Another method to achieve reconfigurability in such distributed structures is with spatial patterning of phase change materials such as vanadium dioxide(VO2). These materials exhibit spatially dependent conductivity between that of an insulator and a metal depending on the local temperature of the medium. By incorporating a micro-heater-array below a film of VO2 conductive circuits can be “drawn” or programmed. However, the conductivity of VO2 in its metallic state is 105 S/m (107 S/m for metals). By incorporating metallic inclusions into the VO2 film, the loss can be reduced to acceptable levels.

In addition to being thermally activated, VO2 can be field activated–that is, by placing an electric field across the material which is beyond some critical switching field, the VO2 can be induced into its metallic state. While the critical field is fixed for the material, the corresponding critical voltage depends upon electrode geometry. In this work, we propose to engineer metallic inclusions with varying radii of curvature and gaps in order to make a settable critical voltage. In such a way a single analog voltage varied from, e.g., 0V to 10V could be used to program the length of an antenna depending on which band radio must operate. We refer to this as a software-defined antenna.

We examined the electric field between two conducting spheres as a way to set the switch point. Theoretical analysis and ANSYS HFSS full-wave electromagnetic simulations were used to quantify the field between the combination of electrode designs. Generally, the peak electric field is inversely proportional to the radius of curvature and gap size. Based upon these results, we designed a programmable antenna with intermittent electrodes of progressively higher critical voltages. In such a way, a low control voltage results in a short antenna (high frequency) and as the control voltage is increased it switches the next (longer) electrode into its metallic state thereby increasing the length of the antenna. We compare the performance of this programmable antenna to a conventional fixed antenna. Gain and input match are shown to be tunable over a wide operating frequency range while the fixed antenna has a single operating band. Future work will aim to develop a prototype based upon this preliminary design.

Principal Investigator: Professor Lin

Other Contributors: Vince Kurtz, Tongjia Zheng

AWaRE REU Researcher: Charles Meyers, University of Notre Dame

Project Description: This REU project aims to develop a team of robotic systems that can accomplish complex team missions even in the face of uncertain and dynamic environments. Applications that motivate this project include, but not limited to, emergency response, future manufacturing systems, and service robots.

In this REU project, we will touch topics not only on hardware/software development but also on theoretical/algorithm design, such as communication-aware coordinated motion planning, task planning through formal methods, a counterexample-guided synthesis which combines logic inference with optimization.

Finding: Modern robotics is an expanding field that has become increasingly popular in a lot of areas, including industry, medicine, and agriculture. There is now an increased interest in coordinating larger groups of robots to parallelize common problems and accomplish more complex tasks. The challenge lies in bridging the gap between individual dynamics and their collective behaviors. In this project, we aim to develop a provably correct framework for designing coordination commands for individual agents such that a global goal can be achieved.

The main task of this REU project is to implement an algorithm that deploys a large group of agents into the desired configuration in a decentralized manner. We model the robots’ configuration as a probability density function (pdf) and compute the velocity field for the agents based on the difference of the current pdf and desired pdf for which the final convergence can be formally proved.

Kernel density estimation is used to estimate the current pdf of the agents. We further parallelize the computation in a way that each agent computes its own local estimation of the current pdf based on its neighbors. This makes it practical when robots have limited sensing, communication and computation capabilities. In order to test this, we simulated Turtlebots using 2D and 3D simulation software to have them converge to distributions and shapes like circles, squares, and triangles. Also, we find that a small number of real Turtlebots are able to converge to a desired distribution.

These results demonstrate that using density feedback control with local estimation to compute velocities can be efficient, accurate, and decentralizable, traits that make it suitable to be introduced into industry.

Principal Investigator: Professor Joshi

Other Contributor: Mark Horeni

AWaRE REU Researcher: Brad King, Texas A&M University

Project Description: Due to tight on-die integration in low-cost, low-power wireless modules, digital and mixed-signal subsystems are often placed very close to each other. Noise coupling from the digital system is often indicative of the computations being performed and thus leaks information to the outside world. We would like to characterize this leakage and see what all can be inferred from power analysis and wireless signal analysis.

Finding: Many Bluetooth chips are vulnerable to wireless attacks because the digital logic and radio transceiver is on the same integrated circuit, causing them to be too close to each other.

The digital circuit performs cryptographic tasks and the radio transceiver broadcasts the signal. The closeness of the two allows information from the digital circuit to leak electromagnetically into the radio transceiver and be transmitted as noise in the Bluetooth signal.

After enough data is collected an algorithm called correlation radio analysis (CRA) can decrypt the keys. Due to the variability of silicon in the fabrication process, it is not likely that a CRA trained on one chip will work on many other chips.

Our research set out to train a convolutional neural network (CNN) using the same data collected for CRA to create a more generalized algorithm. Data was collected wirelessly with a software-defined radio (SDR) while the Bluetooth chip was continuously broadcasting and encrypting a plaintext message. The keys stayed the same while the plaint text changed. An SDR equipped with a well working CNN could be placed in a room discreetly and decrypt Bluetooth signals.

Principal Investigator: Professor Laneman

Other Contributors: D. Scott Null

AWaRE REU Researcher: Gabriella Sanford, University of Dallas

Project Description: The expanding vision for applications of drone swarms has generated significant interest but has also raised numerous technical challenges related to high-speed, low-latency, and reliable communications over drone-to-ground and drone-to-drone wireless links. Widely-deployed commercial radio technologies, e.g., cellular UMTS / LTE as well as WiFi, would seem to offer viable connectivity solutions, but they require characterization of performance tradeoffs, system-level optimizations, and standards enhancements to enable safe drone operations.

To address problems in this space, our team has been developing a low-profile computation and communications platform for drones and collaborating with drone control and software engineering researchers to conduct preliminary flight tests and data collections. The platform consists of a Raspberry Pi embedded computer with a built-in WiFi device as well as a USB UMTS / LTE device that can easily be mounted on a drone. Positioning is obtained for now via a GPS board, or “hat,” for the Raspberry Pi, but it can instead be obtained from the drone itself. Software is being developed to allow the onboard computer to act as a communications manager for the wireless interfaces, allowing the drone to monitor the performance of each wireless interface and controlling the types and frequencies of messages broadcast on each. To prototype these services, we are implementing a distributed detect-and-avoid protocol called Position Intent Broadcast System (PIBS) in collaboration with industry partner InterDigital.

Finding: There is a growing demand for drone swarms, given that they have a wide variety of applications, including search and rescue, surveillance, military operations, and scientific data collection. This increased interest in drone swarms has pushed technical challenges such as collision avoidance and drone communications over wireless links to the forefront of drone-related research. Currently, the platform for drone communication is to have drones receive commands from and send their data to a ground control station, thus requiring one unique telemetry dongle per one drone. Therefore, the number of drones is limited by the number of server USB ports.

However, this project seeks to change the way drones communicate by having them communicate with each other, thereby making drone swarm communications more efficient. We focused on implementing a distributed detect-and-avoid protocol called Position Intent Broadcast System, or PIBS, that was created in collaboration with Interdigital. To do this, we had to work on getting PIBS, which is run on a Raspberry Pi, to pull GPS directly from a SITL (Software in the Loop) drone.

Principal Investigator: Professor Striegel

AWaRE REU Researcher: Alexandra Bejarano, University of Tulsa

Project Description: For a significant portion of the US, a challenge that happens every fall is that of managing the pickup/handling of yard waste due to tree leaves. This is a particularly noteworthy problem for St. Joseph County where the University of Notre Dame is located. A major logistical challenge is how to deploy leaf pickup services to maximize the efficacy of pickup, i.e. it does little good to pick up leaves if there is minimally piled up leaves to pick up. The focus of this REU project will be to explore how we can leverage WiFi to potentially sense the level of leaves still on the trees versus on the ground by virtue of path loss from fixed as well as residential WiFi. This project will explore the extent to which one can fuse passive WiFi sensing from individuals (smartphone apps) and mobile sensors mounted on city vehicles.

Finding: Every fall many places in the US are faced with the management of fallen leaves. In several cities, including the area around the University of Notre Dame, the local municipality is in charge of leaf pick up and disposal to prevent people from simply burning leaves and polluting the air. The issue then lies in knowing the best time to go around and collect leaves. Currently, city vehicles are scheduled to pass through neighborhoods, possibly wasting resources if no/few leaves are there for pick up, or one has to schedule a pick-up. So what if there was a way to automatically know when to go around and collect leaves?

We know that leaves can weaken the strength of WiFi signals. But how reliably could WiFi be used to sense leaves or the lack of leaves on trees? This research was motivated by the following questions: Can one efficiently infer the impact of leaves on WiFi with captures of data packets transmitted over WiFi to possibly determine the most efficient time for leaf pick up? Are packet captures too noisy? Are there enough packets?

Currently, for this research, it captures unencrypted data packets which have been collected with Wireshark, an open-source packet capture and analysis software, on a Raspberry Pi in multiple locations and on multiple WiFi channels. And code has been written to read the packet captures and analyze the information within the files, specifically the access points, signal strength, and a number of packets.

Principal Investigator: Professor Chisum

AWaRE REU Researcher: Tristen Lewandowski, University of Notre Dame

Project Description: As the frequency of distributed microwave circuits corresponds to their physical geometry, there is a need for reconfigurable radios and antennas that operate over a wide range of the electromagnetic spectrum. Currently, advanced RF ICs enable extremely wideband operation, but lack corresponding antennas and other distributed structures. One way to achieve reconfigurability in such distributed structures is with spatial patterning of phase change materials such as vanadium dioxide, or VO₂— when cold, it is an insulator; when heated, it is a conductor. However, a disadvantage of the material is that the conductivity of VO₂is 10⁵ S/m, two orders of magnitude less than that of most metals, 10⁷ S/m, meaning that VO₂ alone dissipates much more energy than metal. The goal of this project is reconfigurable distributed circuits using pixelized metal inclusions on a VO₂ layer with a control mechanism of either spatial UV switching or localized thermal switching.

In preliminary measurements, pure VO₂ CPW transmission lines with metallic launches resulted in extremely lossy lines. In a study of loss and isolation, a coplanar waveguide including unit cells of aluminum and VO₂ was varied in VO₂ thickness, unit-cell gap, CPW cross-section (trace width), and unit-cell size. In the transmission line simulations, aluminum launches were included around the unit-cells, and loss and isolation per unit-cell were extracted. Three different simulators were compared (ADS Momentum, ADS TLIN, and Ansys HFSS) in an attempt to show agreement across software and gain confidence in the model and measured values. Simulation results from parametric sweeps showed ADS Momentum (MoM) and ADS TLIN (Circuit) in close agreement, while HFSS predicts similar trends across frequency but with a higher loss. The approach is promising as results show a sufficiently low loss for a distributed circuit of length λ, but it requires care in balancing loss and isolation. High-quality VO₂ films at greater thicknesses are necessary, and unit-cell gaps should be small (≤200nm). Loss can be further reduced by trading off programmable resolution by using larger unit-cells. The next steps of this project involve fabricating the lines and comparing real measurements to simulations.

Finding: Over the course of the AWaRE REU we successfully measured 1D (linear) programmable transmissions line test structures to confirm the loss and isolation of such lines. We developed three independent simulation models all with significant agreement to the measurements. Models include full-wave electromagnetic (Finite element and Method of Moments), distributed circuit modes (transmissions lines), and lumped element circuit models. The circuit models enable rapid design iteration while the electromagnetic simulations enable high-fidelity simulations.

Principal Investigator: Professor Striegel

AWaRE REU Researcher: Spencer Spitz, Colgate University

Project Description: Although WiFi and cellular are nearly everywhere, the performance of said wireless technologies tends to vary dramatically both in time and space.  While many of us have relied on Speedtest or other tools to help assess just how poor the nearby wireless performance tends to be, the challenge is that tools like Speedtest and others tend to be slow and take significant amounts of bandwidth to run.  We have created a recent set of tools called Fast Mobile Network Characterization that lets us do equivalent tests roughly 100x faster and with 10-100x less bandwidth with approaches that both work by actively probing the network and approaches that are completely passive.  While we have largely demonstrated many of the tools in the lab, the next step is to explore how well the tools translate out into the real world.

Finding: In the technology-driven world of today, WiFi is utilized by billions of people across the globe all the time. So, it is crucial that the WiFi is fast and trustworthy, but bad signals and roaming connections create problems for users. How are we supposed to know where to focus our attention to fix the connectivity issues that arise? The undergraduate REU project focused on developing a proposed solution via an application that generates a lightweight “health assessment” of the WiFi signals in the area.

The developed Android application runs in the background to periodically connect to the web and gather network data via well-structured web requests. With implementations for local and online storage, users may view any data collected in the past as well as data collected by other users. This information is then used to make graphs with visualization as well via Google maps that display the variations of WiFi strength in a creative way. The program runs as a service, gathering network statistics to address connectivity concerns and in turn help to make the world more connected.

Principal Investigator: Professor Laneman

AWaRE REU Researcher: Bob Schenck, Valparaiso University

Project Description: One of the things that makes drones so popular is that a user can communicate with the drone remotely. These communications send information, such as position or battery level, and commands, which instruct the drone in its movements. Drones become dependent on these communications in order to maintain a safe environment for use. By adding an onboard computing and wireless platform to the drone, we hope to advance and expand upon the communication framework available to the drone. This will include several wireless interfaces. A 4G LTE cellular connection will be established for the drone to communicate to the cloud and ultimately establish drone-to-ground communications. Alongside that, an infrastructure less Wifi network is configured on the computer for drone-to-drone communications. The onboard computer also acts as a communications manager for the drone’s wireless interfaces. This allows the drone to monitor the strength of each wireless network and controls the messages broadcast on each. These interfaces and the data we collect from them will allow us to better operate drone swarms and begin implementation of detect-avoid protocols. We aim to conduct tests of the network to examine the strength, speed, range, and other parameters to each medium of communication.

Finding: Our goal for the AWaRE REU program was to configure a flexible wireless communications manager with multiple interfaces (LTE, WiFi) that would enable both drone-to-ground and drone-to-drone communications experiments. The REU student investigated options for the LTE interface on a Raspberry Pi 3B embedded computer, and prototyped a solution based upon a Verizon LTE USB dongle for drone-to-ground communication in collaboration with a staff software engineer. The team also established a UDP broadcast network over WiFi for drone-to-drone communications. Preliminary measurements were collected on signal quality and packet latency as a function of range. We are expanding upon these measurements in order to prototype a distributed detect-and-avoid protocol in collaboration with our industry partner, InterDigital, and planning to publish the results next Spring. More generally, the software engineer will continue to evolve the control messaging and data streaming software architecture to set the stage for follow-on REU projects.

Principal Investigator: Professor Lin

AWaRE REU Researcher: Marina Malone, Loyola University

Project Description: Intelligent physical systems capable of interpreting, planning, and executing high-level tasks requires reasoning over discrete actions and continuous motions, and that poses unique computational challenges. A recent trend in solving these challenges is integrated task and motion planning (ITMP), which proposes to synthesize both discrete task plans and continuous motion trajectories for mobile robots simultaneously. Successful intelligent autonomous robots are able to perform ITMP in uncertain environments, and that requires the ability to observe and act upon its environment to develop a plan towards accomplishing some task. Our basic idea is to use a camera for perception and integrating perception in mission and motion planning. This REU project is mainly about developing a demonstration in unmanned ground vehicle robots by combining perception and learning models with ITMP. We aim to develop a control system in which a robot receives a high-level specification on the continuous robot and object state spaces, but with unreliable prior knowledge.

Finding: In this project, the undergraduate student developed a grasping motion planning for unreliable object position using Kalman Filter, Linear Quadratic Regulator (LQR), and our novel ITMP algorithm, iterative deepening Signal Temporal Logic (idSTL). We use a monocular camera and OpenCV toolbox to estimate the object pose. The novelty on these control systems is that we integrate perception, planning, and control; thus, we can gradually increase the uncertainty of the object’s location by giving unreliable locations of the object to the robot and later obstructing it from the robot’s initial view. This would enable us to evaluate the effectiveness of the robot planning and action taking to find more information about the object it is tasked to find.