Description

Despite the presence of wireless connectivity in most terrestrial scenarios, there are still many extreme environments that cannot be covered by existing wireless networking techniques, including underground, underwater, and confined spaces (tunnels, pipelines, and indoor environments with no network infrastructures). Wireless networks in such environments can enable various highly desired applications in the environmental, industrial, homeland security, law enforcement, and military fields. However, existing wireless techniques do not work in the extreme environments. First, the widely-used solutions based on electromagnetic (EM) waves have extremely small range in the aforementioned environments due to the absence of line-of-sight paths in the air and the high material absorption when penetrating the lossy medium. Second, the acoustic wave-based solution that has been used for under ocean communications suffers strong and dynamic multipath fading in shallow and complex waters, such as rivers and lakes. More importantly, acoustic wave-based solution only works in water medium but do not work if the target environments are complex combinations of air, water, soil, or concrete. Third, although the magnetic induction (MI)-based solution has better penetration capability in most extreme environments, it requires huge antennas to effectively generate the MI signal for reasonable distance, otherwise, the range is still too small for practical applications. Moreover, due to such limitations, current MI technique is only used for point-to-point communications and no MI networking solution has been systematically proposed.

    Intellectual Merit

    In this project, we propose to investigate a new networking paradigm, Metamaterial-inspired Networking (MetaNet), which utilizes our recently developed communication technique, Metamaterial-enhanced Magnetic Induction (M2I), to wirelessly internetwork portable (or even smaller) devices in extreme environments. By equipping each wireless device with a software-defined micro-coil-antenna array (i.e., a smart metamaterial layer), it is possible to achieve reasonable communication range (tens of meters with pocket-sized devices) in various hostile and complex environments, which lay the foundation of MI networking for practical applications. Moreover, since M$^2$I significantly enhances the magnetic coupling among the wireless devices as well as the conductive objects in the surrounding environment, the much closer interactions among all the nodes as well as the environment create both opportunity and risk for network design from physical layer to application layer. The objective of this project is to explore for the first time the fundamentals of metamaterial-inspired networking in various extreme environments through a closed-loop combination of mathematical modeling, simulations, and experimental evaluation. This proposed plan is based on four core intertwined research tasks, i.e. (i) physical layer solutions based on channel analysis of M2I communications in various environments; (ii) environment-aware and cross-layer network control techniques; (iii) network topology-discovery and localization algorithms; and (iv) prototyping and performance evaluation through a MetaNet testbed based on software-defined radios and reconfigurable M2I antenna arrays, as well as a cross layer simulator based on Comsol Multiphysics and NS-3.

    Broader Impact

    MetaNet will generate significant impacts by providing a new networking platform to establish wireless connection in extreme environments. It will eliminate the environmental barriers of various applications, ranging from environmental sustainability, homeland security, to military and defense automation, among others. Education will be integrated with the research in four thrusts. First, a distance education program focusing on everywhere wireless networking will be established to provide both knowledge and experimental experience to online students worldwide, with emphasis on underrepresented communities. Second, a new graduate level course and a senior level capstone course on MetaNet will be developed. The lecture materials will be developed to a new textbook. Third, a tech summer camp on "wireless in bizarre environments" will be organized to reach out K-12 students. Fourth, scientific results will be disseminated at conferences, journals and magazines in the field. Technology transfer activities, including patenting of the proposed concept, will be pursued.


Personnel

    PI: Dr. Zhi Sun

    Ph.D. student: Dr. Hongzhi Guo (Graduated in 2017)

    Ph.D. student: Dr. Xin Tan (Graduated in 2018)

    Ph.D. student: Zhangyu Li (2017-present)

    Ph.D. student: Haochen Hu (2017-present)

    M.S. student: Soham Desai (2018-2019)

    M.S. student: Vaishnendr D. Sudev (2018-2019)


Collaborators

    Dr. Pu Wang, Department of Computer Science, University of North Carolina at Charlotte (in the area of underwater communication and networking).

    Dr. Lu Su, Department of Computer Science and Engineering, University at Buffalo (in the area of wireless sensing).

    Dr. Tarunraj Singh, Department of Mechanical and Aerospace Engineering, University at Buffalo (in the area of smart agriculture).

    Dr. Erasmus Oware, Department of Geology, University at Buffalo (in the area of smart agriculture).


Synopsis

    We will first lay our foundation on the fundamental physics and then move up layers of the networking protocol stack with the following core investigations:

    Task 1. MetaNet channel modeling, active M2I transceiver design, and physical-layer solutions: We will first model the M2I channel in various types of complex environments, which can characterize important tradeoffs, including the choice of the transmission frequency, transmission power, bandwidth, and M2I antenna array size. Then we will focus on develiping the active M2I transceivers where active power is injected into the M2I antenna array to effectively compansite the critical metamaterial loss. The large wireless communication range in various extreme environments that was predicted in theory can be achieved. After that, based on the channel analysis and , we will develop a new software-defined and OFDM-based adaptive physical-layer technique to overcome the dynamic and freqeuncy-selective fading challenges brought by the M2I channel in extreme environments. Accordingly, we will also add the software-defined capability to the M2I antenna design, making the metamaterial sphere smarter, which will leverages the results from our current research on reconfigurable antenna array.

    Task 2. Environment-aware and cross-layer networking techniques: Based on the physical-layer analysis, active M2I transceiver design, and protocols, we will develop the environment-aware and cross-layer (i.e., PHY-Link-Network-layer) network control algorithms to establish reliable and efficient wireless data transmission among distributed M2I devices in different extreme environments. We will first explore the unique MetaNet fast-multihop-relay technique, which utilizes the resonant magnetic coupling between neighbor nodes to complete a multihop transmission only using one time slot. Then we will design network control algorithms to optimally allocate the spectrum, time, and power resources in MetaNet, to take the maximum advantage of the positive influence from the closely coupled M2I neighbors as well as the surrounding environments while mitigating the negative impacts. In additon, we will explore the possible hybrid signal propagtion solutions, such as hybrid M2I and accoustic, to address unique chanllenges in specific environment, such as underwater.

    Task 3. Network topology-discovery, localization, and wireless sensing algorithms: The proposed MetaNet functionalities highly rely on the knowledge of the network topology and devices’ positions. Therefore, we will first investigate the MetaNet topology-discovery technique that senses and learns the surrounding environment to acquire the network topology graphs at all frequency bands. We will also explore the M2I environment-aware localization mechanism that do not rely on any infrastructures or priori knowledge while can still accurately derive the devices’ position. Due to the closely coupling of the above two problems, we will design a jointly algorithm for network topology-discovery and localization based on MetaNet channel models and the kernel-based learning method. Moreover, since the environment-awareness can be achieved in the aformentioned process, we will investigate the possibility to wirelessly sense the environment using MetaNet, for IoT applications in extreme environments, which aims to facilitates novel applictions, such as smart agriculture and complex water exploration.

    Task 4. Prototyping and performance evaluation: We will showcase a fully functional MetaNet wireless network using in-house developed testbeds through real-world experiments. The testbed will consist of USRP X310 software-defined-radio platforms, RF daughterboards, 3D-printed reconfigurable M2I antenna arrays, negative resistance circuit based on operatinal amplifier, and computers and micro-controllers. The experiments will be conducted in three major environments, including indoor, underwater, and underground. In addition, to reduce the experiment cost and time as well as to validate the performance of large scale network, we will also develop a multi-layer simulator based on Comsol, NS-3, and home-developed codes, which can simulate MetaNet’s performance from EM field level to network level.


Outcomes

    Intellectual Merit Outcomes

      Active M2I Transceiver Design

      We have proposed and investigate to utilize active metamaterial units in the M2I communication system, which can close the performance gap between the theoretical analysis and the practical design. Specifically, our preliminary work shows that M2I system is theoretically feasible to increase the signal strength of the original MI system by up to 30 dB (three orders of magnitudes). However, through the real-world design and implementation, we found the implemented M2I communication system only achieves about 8 dB gain in signal strength over the original MI system due to the metamaterial loss. To close the gap, we designed a configurable active circuit-based coil array to form the metamaterial sphere that encloses the MI antenna. The theoretical model of the proposed active M2I system is rigorously developed. Based on the model, the spherical coil array structure of the active M2I transceiver is optimized to maximize the M2I signal strength given the same input power. Through analytical deduction and COMSOL simulations, we showed that the developed active M2I system achieves significant power gain and improvement in communication range compared with the passive M2I system and the original MI system. The developed active metamaterial spherical coil array will serve as the fundamental component of the smart and software-defined M2I transceiver in MetaNet.

      Related publication:

      Z. Li and Z. Sun, "Antenna System Optimization for Active Metamaterial-enhanced Magnetic Induction Communications", in Proc. European Conference on Antenna and Propagation (EuCAP 2019), Krakow, Poland, 2019.

      Active M2I Transceiver Implementation

      We complete the design and development of the smart and software-defined M2I antennas, and named the antenna system as “Reconfigurable Meta-sphere for M2I Communications”. We designed and prototyped the optimal reconfigurable meta-sphere for the M2I communication system, which can approach the theoretical performance bound predicted in the ideal M2I system. In particular, although the ideal M2I metamaterial antenna can theoretically enhance the MI signal strength by up to 30 dB, none of the practical M2I systems can achieve that enhancement due to the difficulty of canceling the metamaterial loss and optimized designing the metamaterial structure. To this end, we proposed to use a reconfigurable active circuit-based coil array to form the metamaterial sphere, which close the gap between the the oretical performance prediction and the practical system achievement. More importantly, we implemented the design and validated the proposed system in real world. In the design, we connect each coil unit on the meta-sphere with a negative differential resistance circuit to cancel the meta-loss. By considering the power consumption's brought by the active elements, the meta-sphere structure is optimized designed to maximize the receiving signal strength given the same input power. In addition, the overall system matching is derived after applying the optimal meta-sphere design to the original MI system. The optimal design is not only validated through COMSOL simulations but also by the real-world experiments using a M2I meta-sphere prototype. The experiment results show that the developed reconfigurable M2I meta-sphere system achieves significant improvement in terms of communication range, compared with the previous developed M2I systems as well as the original MI system. We show the system achieves 18dB overall enhancement of MI signal strength in underground environment.

      Related publication:

      Z. Li and Z. Sun, "Optimal Active and Reconfigurable Meta-sphere Design for Metamaterial-enhanced Magnetic Induction Communications", submitted for journal publication, 2021.


      Full-duplex Relay in MetaNets

      We developed the full-duplex M2I communication technique to realize the envisioned fast-multihop-relay functionality in MetaNet. In particular, the theoretical analysis and electromagnetic simulation are first provided to prove the feasibility. Then, a medium access control protocol is proposed to avoid collisions. Finally, the capacity and delay of the full-duplex M2I network are derived to show the advantage of the new networking paradigm. We showed that in a full-duplex M2I network, the distance between the source and destination can be arbitrarily long and the end-to-end delay can be as short as a single hop delay. As a result, each node in such network can reach any other node by one hop, which can greatly enhance the network robustness and efficiency. It is important for timely transmission of emergent information or real-time control signals. We will use the developed full-duplex M2I technique in the next step networking control algorithm design. More details are given in the next sections.

      Related publication:

      H. Guo and Z. Sun, "Full-duplex Metamaterial-enabled Magnetic Induction Networks in Extreme Environments", in Proc. IEEE Infocom 2018, main conference, Honolulu, USA, April 2018.


      M2I assisted acoustic cooperative MIMO

      To establish reliable and long-range underwater link in MetaNet, we develop the M2I assisted acoustic cooperative MIMO that takes advantages of both M2I and acoustic underwater communication techniques. While we believe acoustic system is still the solution for long-range underwater communication between underwater nodes and surface station, the acoustic channel is not reliable due to the dynamic underwater conditions. Although MIMO can seek the spatial diversity to enhance the reliability, it was impossible for underwater communications before, due to two reasons: (i) centralized MIMO require one wireless underwater node carry multiple antennas that are sufficiently separated, which is impossible for the size of most underwater sensors or robots; and (ii) the distributed and cooperative MIMO that relies on the collaboration between independent underwater nodes requires accurate synchronization that is impossible for underwater acoustic channel. To this end, we developed the hybrid system that utilizes the M2I to synchronize the fish in a swarm and cooperatively form a acoustic beam towards the surface station. Through theoretical analysis, computer simulations, and in-lab experiments, we have showcased the effectiveness of this very novel idea.

      Related publication:

      S. Desai, V. Sudev, X. Tan, P. Wang, and Z. Sun, "Enabling Underwater Acoustic Cooperative MIMO Systems by Metamaterial-Enhanced Magnetic Induction", in Proc. IEEE Wireless Communications and Networking Conference (WCNC 2019), Marrakech, Morocco, April 2019.

      H. Guo, Z. Sun, and P. Wang, "On Reliability of Underwater Magnetic Induction Communications with Tri-Axis Coils", in Proc. IEEE ICC 2019, Shanghai, China, May 2019.

      Z. Li, S. Desai, V. D Sudev, P. Wang, J. Han, and Z. Sun, "Underwater Cooperative MIMO Communications using Hybrid Acoustic and Magnetic Induction Technique", Computer Networks Journal (Elsevier), Vol. 173, May 2020.


      M2I-based Environment-aware Localization Technique

      We develop an environment-aware localization strategy for M2I-based wireless networks in complex environments with arbitrary number of conductive objects that have arbitrary sizes and shapes. Specifically, the influence of conductive objects on the MI network in indoor environment is first investigated. The effects of the conductive objects with different dimensions, directions, and positions are analytically captured. Then a joint device localization and conductive-object tomography algorithm is developed to estimate the position of each wireless devices as well as distribution of objects. The localization technique is implemented by a series of experiments in the complex environments, including a general indoor environment, an environment with large metallic objects, a metallic pipe-like environment, and a simulated underwater environment (in-lab tank). Moreover, a hybrid solution using both MI and inertial sensors is designed and implemented. By using the inertial sensors to measure the orientation of each MI sensor node, not only the inter-node distance can be estimated, but also the relative direction of the nodes can be determined.

      Related publication:

      X. Tan, Z. Sun, P. wang, and Y. Sun, "Environment-Aware Localization for Wireless Sensor Networks using Magnetic Induction", Ad Hoc Networks Journal (Elsevier), Vol. 98, March 2020.


      MetaNet-based Soil Moisture Sensing System

      To achieve real time high precision soil moisture sensing in a large range underground environment, we developed a MetaNet-based soil moisture sensing system. Unlike existing work treat the testing area as homogeneous medium, we consider the soil moisture sensing in real inhomogeneous environment. We divide the inhomogeneous sensing area into small cubic units. The soil moisture in each unit is considered to be homogeneous. M2I wireless nodes are buried along the perimeter of the sensing area. Each time only one pair of MI sensors establish a communication link, in a short time duration a group of communication links are established to provide enough data for real time sensing. The high penetrability intrinsic of M2I links ensure that any two pair M2I node can cover a large sensing range. Therefore, the nods density for the proposed sensing scheme is much lower than traditional GPR-based methods. The channel model of underground M2I link in the cubic-unit-based soil environment is rigorously deducted. We developed the phase delay based moisture estimation scheme using the above MetaNet-based sensing system. Through full-wave simulation in COMSOL, the data are collected for estimation. Our estimation results show that in the underground environment with 40 meters range and 30 meters depth, the proposed MetaNet-based soil moisture sensing system can achieve very accurate moisture sensing result.

      Based on the theoretical design of the M2I large range soil moisture sensing system, we developed a soft-ware defined indoor soil moisture sensing test-bed. The test-bed is based on the USRP N210 platform and is equipped with LFTX and LFRX daughter-boards that can operate from 0 to 50 MHz. The proposed soil moisture sensing system utilizes the channel state information (CSI) of the underground M2I links in the MetaNet to estimate the water content in the inhomogeneous soil medium. Each underground M2I link is established between two M2I communication modules. An in-lab sand box is used as the test soil environment, which is open on the top and has a water pump at the bottom. The volumetric water content (VWC) in the test environment is controlled by adding water from the top and pumping water out from the bottom of the tank. To evaluate the performance of the soil moisture distribution estimation in an inhomogeneous underground environment, the test soil environment is divided into a certain number of cubic soil units with the same size. The VWC in each unit is approximately considered to be homogeneous. The M2I communication modules are activated sequentially so that the CSI of all the M2I links in the MetaNet can be measured in a short period. We utilize the underground M2I channel model to estimate the soil VWC based on the signal propagation phase delay from the CSI reading. As shown in the table in the third picture below, the sensing results are very accurate in an inhomogeneous soil environment.

      Related publication:

      Z. Li, Z. Sun, T. Singh, and E. Oware, "Large Range Soil Moisture Sensing for Inhomogeneous Environment using Magnetic Induction networks", in Proc. IEEE Globecom 2019, Waikoloa, USA, December 2019.

      Z. Li, Z. Sun, T. Singh, and E. Oware, "Real-time and High-precision Soil Moisture Sensing for Inhomogeneous Environment using Magnetic Induction networks", submitted for journal publication, 2021.


      Underwater Activity/motion Recognition using Hybrid M2I-accoustic Networks

      After developed the hybrid M2I-accoustic underwater network, we used the network in the application of underwater activity/motion recognition in the third year. In the hybrid M2I-accoustic underwater sensing system, the M2I links are used to provide low-latency links between distributed acoustic transponders, while the acoustic links to and from the distributed acoustic transponders are used to sense the underwater objects and their activities. The sensing system only uses the hybrid M2I-accoustic communication infrastructure, no expensive underwater cameras are needed, which is different from the widely used image based underwater classification techniques. We have been focused on two aspects of the proposed sensing system: (1) the underwater sensing mechanism based on distributed acoustic wireless networks, and (2) the theoretical guidelines and performance bound analysis in wireless network-based sensing system. In particular,

      • We developed an underwater motion recognition mechanism in the inhomogeneous deep-sea water using acoustic wireless networks. To accurately extract velocities of target body features, we first derive doppler frequency shift (DFS) coefficients that can be utilized for Velocities of target body (VTB) estimation when signals propagate deviously. Secondly, we propose a dynamic self-refining (DSR) optimization algorithm with acoustic wireless networks that consist of multiple transmitter receiver links to estimate the VTB. Those VTB features can be utilized to train the convolutional neural networks (CNN). Through the simulation, estimated VTB features are evaluated and the testing recognition results validate that our proposed underwater motion recognition mechanism is able to achieve high classification accuracy.
      • Meanwhile, we realized that although machine learning algorithms can achieve high recognition accuracy by training with plenty of data and adjusting architectures, existing works lack a theoretical performance bound analysis that can capture the effects of the influencing factors in wireless networks, such as the underwater physical environments and settings of the hybrid M2I-acoustic devices. To this end, we investigated the wireless sensing performance upper bound based on the maximum likelihood principle. The performance upper bound is obtained as a function of influencing factors in wireless sensing networks. Through extensive experiments (as shown in the following pictures) and theoretically analysis, we evaluated the influences of different influencing factors and use the derived performance upper bound to optimize the deployment of the wireless networks to improve the wireless sensing accuracy.

      Related publication:

      H. Hu, Z. Sun, and L. Su, " On Influencing Factors in Human Activity Recognition using Wireless Networks", in Proc. IEEE Globecom 2019, Waikoloa, USA, December 2019.

      H. Hu, Z. Sun, and L. Su, "Underwater Motion and Activity Recognition using Acoustic Wireless Networks", in Proc. IEEE ICC 2020, Dublin, Ireland, June 2020.

      H. Hu, Z. Sun, and L. Su, "On Upper Bound of Human Activity Recognition using Wireless Sensing", submitted for journal publication, 2021.

      H. Hu, Z. Sun, and L. Su, "Distributed Wireless Motion and Activity Recognition using Underwater Acoustic Networks", submitted for journal publication, 2021.


      M2I-based Beamforming for Wireless Power Transfer

      Since the energy source is very limited in the underground and underwater environments, it is essential to develop efficient wireless power transfer technique to charge the devices. Maximizing the power transfer efficiency (PTE) is one of the most crucial problems. The research community has been using MI beamforming to maximize the PTE for the near field MIMO WPT systems. However, conventional magnetic beamforming techniques in the literature require precise magnetic channel information, which is very difficult to obtain in extreme environments like underground and underwater. Therefore, we utilized the M2I and reconfigurable metasurface technique from this project to develop a novel MI beamforming scheme in the MIMO WPT system. The reconfigurable metasurface is consist of an active metamaterial layer, which is settled above the power source. Similar to the active M2I meta-sphere design, the non-foster active circuit is used to obtain negative resistance and capacitor in each coil unit on the metamaterial layer. Instead of controlling the source currents or voltages, the reconfigurable metasurface can achieve near field beamforming only by varying the capacitor and resistance in specific coil array units. The MIMO WPT transmitter estimates the receiver’s position by just measuring its own current. Then the MI beamforming is formulated as a discrete optimization problem and solved by the Simulate Anneal (SA) method. Through the analytical deduction and COMSOL simulations, we showed that the proposed beamforming scheme can achieve approximately two times PTE compared with the conventional MI beamforming techniques within a 40 cm charging distance.

      Related publication:

      Z. Li and Z. Sun, "Enabling Magnetic Beamforming in MIMO Wireless Power Transfer Using Reconfigurable Metasurface", in Proc. IEEE Globecom 2020, online, December 2020.


      More over, recently we realized the proposed MIMO WPT transfer scheme based on a reconfigurable metasurface, as shwon in the following picture. Experimental results show that the MIMO WPT system using the proposed beamforming scheme can improve the overall system PTE and enlarge the effective power region significantly over conventional magnetic beamforming scheme.

      Related publication:

      Z. Li and Z. Sun, "Reconfigurable Metasurface Enabled Magnetic Beamforming for Wireless Power Transfer", submitted for journal publication, 2021.


    Broader Impact Outcomes

      Educational Activity

      PI Sun has incorporated the latest results in terms of wireless underground and underwater communication to his course "EE441/541: Wireless Communications in LTE, mobile TV, and Extreme Environments", in the Department of Electrical Engineering at the University at Buffalo, to both senior undergraduate and graduate students.

      Moreover, PI Sun has developed a new senior graduate level course "EE605: From LTE to 5G and Cyber Physical System" in fall 2018, where the MetaNet system will be an important component in the Cyber Physical System in extreme environments.

      In addition, PI Sun has incorporated the latest results in terms of wireless underground and underwater communication to the senior level capstone course course "EE494: Senior Design Implementation". Senior undergraduate students can use the recently developed underground and underwater testbeds in PI Sun's lab to design their own system for emerging applications.

      The remote education testbed for MetaNet is being built to enable the distance education program focusing on everywhere wireless networking.

      PI Sun is currently writting a book titled "Magnetic Communications: Theory and Techniques", with the Cambridge University Press. Two chapters of the book will cover the theory and implementation developed in the MetaNet project.

      Outreach and Other Broader Impact Outcomes

      PI Sun and his group have been hosting students from local high schools in Buffalo in the "Science Exploration Day" every year starting from 2016. The lab is opened for short courses and demonstrations.

      PI Sun is leading a multi-disciplinary research team (from Soil and Crop Sciences Center at Cornell University and multiple departments at University at Buffalo) to investigate the next generation underground soil sensing system for future smart agriculture applications. We propose to develop a disruptive technology that enables high spatial and temporal monitoring of dynamic soil properties by addressing the key constraints of existing solutions, including: requiring direct contact with soil, lacking the sensing capability of dynamic properties of plant-available, metabolizable and movable fractions with high sensitivity, and last but not the least, the very limited sensing range.

      PI Sun is participating in a multi-disciplinary research team (from multipe departments at University at Buffalo, Rochester Institute of Technology, and Sandia National Laboratories) to develop the self-sustainable energy cloud for situational awareness in resilient solar energy microgrids. PI Sun's focus lies in the robust and reliable wireless communication networks that enables continuous real-time monitoring and communication from sensors in extrem conditions and environments.