Highlights
SMAC-FIRE: Closed-Loop Sensing, Modeling and Communications for WildFIRE
PIs: Lee Swindlehurst, Ahmed Eltawil, Tirtha Banerjee, Zak Kassas, Janice Coen (UCAR), and Hamid Jafarkhani
Increases in temperatures and drought duration and intensity due to climate change, together with the expansion of wildlife-urban interfaces, has dramatically increased the frequency and intensity of forest fires, and has had devastating effects on lives, property, and the environment. To address this challenge, this project's goal is to design a network of airborne drones and wireless sensors that can aid in initial wildfire localization and mapping, near-term prediction of fire progression, and providing communications support for firefighting personnel on the ground. Two key aspects differentiate the system from prior work: (1) It leverages and subsequently updates detailed three-dimensional models of the environment, including the effects of fuel type and moisture state, terrain, and atmospheric/wind conditions, in order to provide the most timely and accurate predictions of fire behavior possible, and (2) It adapts to hazardous and rapidly changing conditions, optimally balancing the need for wide-area coverage and maintaining communication links with personnel in remote locations. The science and engineering developed under this project can be adapted to many applications beyond wildfires including structural fires in urban and suburban settings, natural or man-made emergencies involving radiation or airborne chemical leaks, "dirty bombs" that release chemical or biological agents, or tracking highly localized atmospheric conditions surrounding imminent or on-going extreme weather events.
The system developed under this project will enable more rapid localization and situational awareness of wildfires at their earliest stages, better predictions of both local, near-term and event-scale behavior, better situational awareness and coordination of personnel and resources, and increased safety for fire fighters on the ground. Models ranging from simple algebraic relationships based on wind velocity to more complex time-dependent coupled fluid dynamics-fire physics models will be used to anticipate fire behavior. These models are hampered by stochastic processes such as the lofting of burning embers to ignite new fires, that cause errors to grow rapidly with time. This project is focused on closing the loop using sensor data provided by airborne drones and ground-based sensors (GBS). The models inform the sensing by anticipating rapid growth of problematic phenomena, and the subsequent sensing updates the models, providing local wind and spot fire locations. Closing this loop as quickly as possible is critical to mitigating the fire's impact. The system we propose integrates advanced fire modeling tools with mobile drones, wireless GBS, and high-level human interaction for both the initial attack of a wildfire event and subsequent on-going support.
Distributed and Quantized Kernel-Based Learning over Interconnected Sensing Systems
PIs: Yanning Shen and Hamid Jafarkhani
Kernel-based learning is widely used for nonlinear function learning, which is a general task in various machine learning problems, e.g., classification and regression. This leads to its wide application in pattern recognition and data analysis in many challenging tasks, including but not limited to time series prediction in various interconnected sensing systems such as sensor and IoT networks. For example, sensors measure the temperature, humidity, pressure, or other physical phenomena to predict future measures, and cameras take pictures, or videos to recognize an object. In many applications, multiple access nodes collect and/or disseminate information over a certain geographical area of interest. In addition, sensing systems include many computationally capable devices like smartphones, UAVs, cars, and so on. In such distributed networks, both data and computational power are distributed. Transmitting the collected data back to a central entity for processing is not desired. Also, it is impossible to transmit the massive amount of collected data in real-time over networks. In addition, in many applications, there are valid security and privacy concerns about transmitting personal data, for example in medical and finance applications. Therefore, the proposal aims to design distributed and quantized learning algorithms that do not transmit the collected data over networks.
We design online distributed and quantized kernel-based learning algorithms that calculate some "local" updates and communicate the corresponding "updates" to their neighbors such that collectively the network can learn a "global" model. This is done without transmitting the collected data over sensing systems with static and dynamic network architectures. We present different distributed and quantized function learning algorithms and study their performance including their convergence and regret analysis. Our algorithms will be designed for different network structures while accounting for network delays and dynamics. We also design adaptive distributed and quantized algorithms and study their performance and regret analysis. In addition, we study the optimal network resource allocation in these scenarios and the corresponding trade-off between computation accuracy and network resources. Our goal is to design robust distributed and quantized kernel learning algorithms over distributed sensing systems that only need to communicate with neighboring nodes and are less sensitive to network characteristics, like network topology and delays.
Flexible Coding for Distributed Storage and Computing
PIs: Zhiying Wang, Hamid Jafarkhani, and Syed Jafar
Due to the high demand for accessing and storing a large amount of data, distributed systems consisting of hundreds of thousands of devices are widely used. In distributed systems, failures are quite common and hard to predict. Depending on the system parameters, it is desirable to allow adaptive choices of code constructions or computation schemes that provide desirable cost functions, such as storage size, computational complexity, and latency. In this project, the constructions and schemes for flexible coding in distributed storage and computing are investigated.
First, facing the fact that the failures are unpredictable in a distributed system, a framework for flexible storage codes to achieve the optimal latency of accessing information is proposed. Instead of accessing a fixed number of nodes as in a conventional code, a flexible code allows one to recover the entire information from a flexible number of storage nodes, and reduce the accessing latency. Constructions for different storage scenarios are proposed, including storage with minimum redundancy, with minimum repair bandwidth, with local repair, and with correction for mixed types of errors. Second, flexible coding in distributed matrix multiplication for failure tolerance is proposed to reduce the computing load. A master node obtains the computation results from the available servers, and should be able to recover the matrix product. The number of available servers is unknown a priori. Code constructions are established that can efficiently make use of the computation results from all available servers. Given the storage capacity of the servers, the computation load is optimized.
Simplified Implementation and Control of a Reflective Intelligent Surface (RIS)
PIs: Ender Ayanoglu, Lee Swindlehurst, and Filippo Capolino
Professors Ender Ayanoglu, Lee Swindlehurst, and Filippo Capolino are working on a project that promises to simplify the implementation and control of a reflective intelligent surface (RIS). They carry out this work as part of the NSF grant 2030029. RISs are a very active research area. They are made up of reflective elements which can change the phase of an impingent electromagnetic wave and thus can work as a reflector, or mirror, for such waves. They can be constructed by means of a technology known as metasurfaces, which can be implemented by tiling a substrate surface with reflective elements. A common technique to implement such reflective elements is using a varactor diode, which is a reverse-biased diode that acts as a variable capacitor whose value can be altered by the bias voltage applied. As such, the element causes a phase and amplitude change in its electromagnetic reflection coefficient. This results in the change of direction of the impingent wave. By optimizing the individual phase and magnitude changes, the received power at a particular location, for example, the location of an individual user, can be maximized. RISs have several potential uses, the most common of which is the redirection of a transmitted wave towards a user in case the user is electromagnetically blocked by natural or man-made objects, such as a large building.
RISs require different bias voltages for each of the RIS elements. This is typically achieved by using different wires or printed circuits. As an RIS can have many elements, say of the order of hundreds or even thousands, this conventional control can become messy and difficult. The work that Professors Ayanoglu, Swindlehurst, and Capolino is conducting is based on replacing the electrical control with one based on a standing wave. In this approach, transmission lines along the RIS surface are formed and a standing wave is generated on each transmission line. The RIS picks up the bias voltage via the value of the standing wave in its location. Thus, electrical control is eliminated and the control of the RIS is greatly simplified. The project is described in a paper by the three researchers published in the August 2022 issue of the IEEE Wireless Communications Magazine.
Available Software
- Multiscale Image Quality Estimator (MIQE) Software. Link
- Multiscale Video Quality Estimator (MVQE) Software. Link