Nanopower Analog Frontends for Cyber-Physical Systems



PhD Defense:   Kenji   Aono

Cyber-Physical System
US DoT ABC University Transportation Center
Cyber-Physical & IoT (Aono 2016)
Challenges

Sensing events of interest over long periods

Recording rare, sparse events

Filter & Data-logging

Wiring is impractical, battery needs to match lifespan

100 mAh battery for 20 years

\[\frac{100 \,\mathrm{mAh}}{175,\!440\,\mathrm{h}} \approx 570\,\mathrm{nA}\]

Nanopower

Frontends

Contributions of Dissertation
Three approaches towards nanopower cyber-physical frontends

First to exploit jump resonance in analog filters to improve acoustic recognition

Investigate a unified filtering and measurement device using analog construct

Demonstrate quasi-self-powered platform for long-term deployment

Case Study Biquad Filter

\[I_\mathrm{Out} = (V_\mathrm{+} - V_\mathrm{-})\,g_\mathrm{m},\quad g_\mathrm{m}\propto I_\mathrm{Bias}\]

\[V_\mathrm{Bp} = \dfrac{s\left(\frac{g_\mathrm{m1}C_\mathrm{2}}{g_\mathrm{m2}\,g_\mathrm{m3}}\right)}{s^2\left(\frac{C_\mathrm{1}\,C_\mathrm{2}}{g_\mathrm{m2}\,g_\mathrm{m3}}\right)+s\left(\frac{C_\mathrm{2}}{g_\mathrm{m3}}\right)+1}\]

\[Q = \sqrt{\frac{g_{m2}}{g_{m3}}}\]

\[\omega_0 = \frac{\sqrt{g_{m2} \times g_{m3}}}{C}\]

\[G = \frac{g_{m1}}{g_{m3}}\]

Furth (1996)
Non-linearity in Biquad Filter

Ideal

Saturated

Linear Filter


Jump Resonance
ASIC Implementation
Parameter Value
Technology 0.5 µm CMOS
Supply Voltage 3.3 V
Fitler Frequency 100 Hz — 4 kHz
Transconductance 800 pS — 32 nS
Capacitance 1.28 pF
Input Voltage 100 mV
Power Dissipation 3 nW — 30 nW
Measured Data for Linear Filter
Measured Jump Resonance Aono (2012)
Jump Resonance in Literature
  • Electrical Engineering
    • Control theory to define regions of jumps 1
    • Analog designers try to squash it 2
  • Biological
    • Has been inferred for 70 years 3
    • Recent studies verify its existence 4
    • Benefits are still being investigated 5
1 Scott (1952), 2 Sarpeshkar (2006), 3 Gold (1948), 4 Lukashkin (2007), 5 Guerreiro (2018)
Speaker Recognition Aono (2013)
Software Linear Jump Software
EER PD EER PD EER PD EER PD
2.01 98.75 2.01 95.63 0.66 100 0.76 99.38

Testing on 10 speakers from YOHO, using SVM

Summary of Contribution 1
  • Nanopower
    • Sub-nA supply current
    • Tunable hysteresis
  • Nonlinear
    • Enables high-Q with few active components
    • Encodes frequency trajectory
  • Recognition
    • Using SVM classifier
    • Jump resonance feature matches MFFC+Δ
Contributions of Dissertation
Three approaches towards nanopower cyber-physical frontends

First to exploit jump resonance in analog filters to improve acoustic recognition

Investigate a unified filtering and measurement device using analog construct

Demonstrate quasi-self-powered platform for long-term deployment

Sensing Frontend in Analog Domain

  • Always-on sensing
  • Reduced energy

Huang (2011)
Digital Data-Logging
Univ. Washington (2019)
Impact-Ionized Hot-Electron Injection (IIHEI)

  • Electric field pushes electrons over Si-SiO2 barrier
  • Gate of MOSFET is electrically isolated
  • Charge remains on gate, changing MOSFET response
Linearized IIHEI
  • Use feedback to maintain field potential
  • Enables predictable, linear injection rate

\[I_\mathrm{inj} = \alpha I_\mathrm{S}\exp\left(\frac{\lambda V_\mathrm{sd}}{V_\mathrm{inj}}\right)\exp\left[\frac{-\beta}{\left(V_\mathrm{gd} + \delta\right)^2}\right]\]

Huang (2011)
Linear Injection Response Aono (2018)

Fabricated PFG SoC Aono (2015,2018)
Parameter Value
Technology 0.5 µm CMOS
Supply Voltage 1.8 V
FG Capacitance 100 fF
Power (Initialize) 500 nW
Power (Biasing) 250 nW
Minimum Energy 100 nJ
Maximum Precision 13.4 bits
Filter Enhancement
Elvin (2006), Lajnef (2014)
Proposed Modification

  • Analog filter
  • Subthreshold

Modified Design

Biquad Filter

Linear Injector

Combined

\[V_\mathrm{out} = \dfrac{V_\mathrm{Ref}}{\frac{s}{g_\mathrm{m2}}\left(C_\mathrm{in}+C_\mathrm{FG}\right)+ H(s)}\]

\[\!\!\!\!+ \dfrac{sC_\mathrm{in}V_\mathrm{in}/\!g_\mathrm{m2}}{\frac{s}{g_\mathrm{m2}}\left(C_\mathrm{in}+C_\mathrm{FG}\right) + H(s)}\]

\[H(s) = \dfrac{g_\mathrm{m3}}{sC_\mathrm{H}+g_\mathrm{m3}}\]

Filtering Behavior
Enhanced Sensitivity
Summary of Contribution 2
  • Linearized Injection Sensor
    • Zero downtime, continuous logging
    • Self-powered (signal of interest is energy source)
  • Leverage Contribution 1 to realize filter
    • Modification enables lower input stimuli
    • Nanopower filter and measurement
Contributions of Dissertation
Three approaches towards nanopower cyber-physical frontends

First to exploit jump resonance in analog filters to improve acoustic recognition

Investigate a unified filtering and measurement device using analog construct

Demonstrate quasi-self-powered platform for long-term deployment

Mackinac Bridge

Need to enable communication greater than 100 m

Battery is required, but must be periodic

Quasi-self-powered

  • Frontend of Contribution 2
  • Battery-backed Interface

Long Range Interface
  • Self-powered sensor logs cummulative history
  • Battery-powered interface supply current
  • \[\mathrm{Request} = \frac{I_\mathrm{on}t_\mathrm{on}+I_\mathrm{search}t_\mathrm{search}+I_\mathrm{off}t_\mathrm{off}}{t_\mathrm{on}+t_\mathrm{search}+t_\mathrm{off}}\]

    \[ \qquad= \frac{2.5\,\mathrm{m}\cdot 13.5 + 225\,\mathrm{µ}\cdot 6 + 50\,\mathrm{n}\cdot 300}{13.5+6+300} < 110 \,\mathrm{µ A}\]

    \[\mathrm{No\: request} = \frac{I_\mathrm{search}t_\mathrm{search}+I_\mathrm{off}t_\mathrm{off}}{t_\mathrm{search}+t_\mathrm{off}} < 5 \,\mathrm{µ A}\]

    \[1.2\,\mathrm{Ah} / \left(0.99\!\cdot\!5\,\mathrm{µ} + 0.01\!\cdot\!110\,\mathrm{µ}\right)\!\!\mathrm{A}\!\cdot\!\left(\!\frac{1\,\mathrm{yr}}{8766\,\mathrm{h}}\!\right) \approx 23.5\,\mathrm{years}\]

    System Supply Currents

    Using long-range sub-GHz wireless module from TI

    2016 Prototype
    2017 Prototype

    a) Nanopower Timer,  b) Interface to PFG,  c) Battery Management,  d) RF MCU,  e) Antenna,  f) PFG Module1

    1Lajnef (2008)
    Labor Day Walk
    2017 and 2018 Recordings


    Summary of Contribution 3
    • Powered Wireless Interface
      • Technology-agnostic, leverage latest COTS
      • Long transmission range
    • Self-powered Sensing and Data-logging
      • Enables power-hungry MCU to shutdown
      • Cumulative logging of input stimuli
      • Decades of designed operational lifespan
    Mackinac Bridge: The Mighty Mac

    Sensors installed on steel beam

    Under the Mackinac Bridge. The bridge sways as much as 35 ft and is 155 ft from the water.

    Federal Aviation Administration

    Accelerated taxiway testing in a hangar with model aircraft wheels

    National Renewable Energy Laboratory

    Delamination/cracking from side

    Sensors mounted on blade

    Nantes, France
    Sensor testing at unique European Union facility.
    Representative Pavement Data

    Linear data-logging of strain

    Distributions show shift as damage is induced

    US DoT ABC University Transportation Center

    Piezoelectric cantilevers on deck

    Full-scale Northridge Earthquake

    Acceleration Analysis

    PDF of collected data

    Peaks of distribution exhibit a pattern

    Femur Fixation Tracking

    Smart Mouthguard

    PFG in mouthguard

    Additive printing circuit connections

    Self-powered Wireless

    Different form factors

    Front of nickel-size interface

    Back of nickel-size interface

    Hybrid Radio-Frequency Interface

    Evolution of the quasi-self-powered sensor concept — hybrid RF uses energy scavenging to trigger a battery-powered transmission

    Testing with truck reader

    Data Interpretation
    Analysis of collected data is done by domain experts

    Deployment Sites & Collaborations
    • U.S. National Science Foundation (NSF)
    • U.S. Federal Highway Administration (FHWA)
    • U.S. Federal Aviation Administration (FAA)
    • U.S. Air Force Research Lab (AFRL) Materials and Manufacturing Directorate
    • National Wind Technology Center (NWTC)
    • National Renewable Energy Laboratory (NREL)
    • Michigan Department of Transportation (MDoT)
    • Mackinac Bridge Authority (MBA)
    • Florida Department of Transportation (FDoT)
    • Michigan State University
    • University of Nevada, Reno
    • University of Southern California
    • University of Missouri, Columbia (MU)
    • Institut Français des Sciences et Technologies des Transports, de l’Aménagement et des Réseaux (IFSTTAR)
    • University of Nottingham
    • University of Palermo
    Contributions of Dissertation
    Three approaches towards nanopower cyber-physical frontends

    First to exploit jump resonance in analog filters to improve acoustic recognition

    Investigate a unified filtering and measurement device using analog construct

    Demonstrate quasi-self-powered platform for long-term deployment

    Publications (Archival)

    [0.1] Aono, Chowdhury, Chakrabartty, Ross. "Speaker Identification in Degraded Audio Samples with MFCC and SPARK Feature Fusion," (in preparation).

    [0.2] Hasni*, Aono*, Pochettino, Lajnef, Chakrabartty. "A Robust Fatigue Crack Detection in Wind Turbine Blade: Battery-Free Sensing Approach," (in preparation).

    [0.3] Aono, Mehta, Chakrabartty. "Unified Filtering and Data-Logging Device," IEEE TCAS-I, (in preparation).

    [0.4] Hasni*, Aono*, Pochettino, Lajnef, Chakrabartty. "Long-Term Structural Health Monitoring of Steel Bridges Using Quasi-self-powered Sensors: The Mackinac Bridge Case Study," Structural Health Monitoring, (in preparation).

    [0.5] Alazzawi, Aono, Scheller, Chakrabartty. "Exploiting Self-capacitances for Wireless Power Transfer," IEEE TBioCAS, (submitted).

    [0.6] Zhou, Kondapalli, Aono, Chakrabartty. "Dynamic Signatures for Authenticating IoT Devices Using Self-powered FN Tunneling Timers," IEEE IoT, (submitted).

    [0.7] Aono, Pochettino, Hasni, Lajnef, Chakrabartty. "Quasi-self-powered Piezo-Floating-Gate Sensing Technology for Continuous Monitoring of Large-Scale Bridge," Frontiers Built-Environment, (submitted).

    [0.8] Alavi, Hasni, Jiao, Aono, Lajnef, Chakrabartty. "Advances in Battery-free, Wireless Civil Infrastructure Health Monitoring Integrated with Machine Learning," Transportation Research Board, 2019 (submitted).

    [1] Zhou*, Aono*, Chakrabartty. "A CMOS Timer-Injector Integrated Circuit for Self-Powered Sensing of Time-of-Occurrence," IEEE Journal of Solid-State Circuits, 2018.

    [2] Gangopadhyay, Aono, Mehta, Chakrabartty. "A Coupled Network of Growth Transform Neurons for Spike-Encoded Auditory Feature Extraction," bioRxiv, 2018.

    [3] Hasni, Alavi, Jiao, Lajnef, Chatti, Aono, Chakrabartty. "A New Approach for Damage Detection in Asphalt Concrete Pavements Using Battery-free Wireless Sensors with Non-Constant Injection Rates," Measurement, 2017.

    [4] Borchani*, Aono*, Chakrabartty. "Monitoring of Postoperative Bone Healing Using Smart Trauma-Fixation Device with Integrated Self-powered Piezo-Floating-Gate Sensors," IEEE TBME, 2016.

    [5] Feng*, Aono*, Covassin, Chakrabartty. "Self-powered Monitoring of Repeated Head Impacts Using Time-Dilation Energy Measurement Circuit," IEEE TBioCAS, 2015.

    [6] Aono, Shaga, Chakrabartty. "Exploiting Jump-Resonance Hysteresis in Silicon Auditory Front-ends for Extracting Speaker Discriminative Formant Trajectories," IEEE TBioCAS, 2013.

    [7] Chakrabartty, Feng, Aono. "Noise-shaping Gradient Descent-based Online Adaptation Algorithms for Digital Calibration of Analog Circuits," IEEE TNNLS, 2013.

    [0.9] Fazel, Aono, Chakrabartty. "Evaluation of SPARK Speech Features for Noise-Robust Speaker Verification," IEEE Signal Processing Letters, 2012.

    Publications (Other)

    [0.1] Aono, Mehta, Chakrabartty. "On The Filtering Behavior Observed in Limit-Driven Linear Feedback Injectors," ACM GLSVLSI, 2019 (in preparation).

    [0.2] Hasni, Aono, Lajnef, Chakrabartty. "Continuous Crack Monitoring on Wind Turbine Blades," SPIE, 2019 (in preparation).

    [0.3] Pochettino, Aono, Hasni, Lajnef, Chakrabartty. "Infrastructural Internet-of-things using Quasi-self-powered Structural Health Monitoring Sensors," SHMII-9, 2019 (submitted).

    [0.4] Kondapalli, Pochettino, Aono, Hasni, Lajnef, Chakrabartty. "Embedded H-gage with Self and Quasi-self-powered Sensors for Pavement Monitoring," SHMII-9, 2019 (submitted).

    [0.5] Kondapalli, Zhou, Aono, Chakrabartty. "Self-Powered CMOS Time-Synchronized Temperature Monitoring," IEEE ISCAS, 2019 (submitted).

    [0.6] Pochettino, Kondapalli, Aono, Chakrabartty. "Enabling Long-Term Infrastructure to Vehicular Communication with Hybrid Powered Systems," IEEE ISCAS, 2019 (submitted).

    [1] Hasni, Lajnef, Alavi, Aono, Chakrabartty. "Local-global Damage Identification Approach Using Hybrid Network of Self-powered Sensors," 7WCSCM, 2018.

    [2] Hasni, Lajnef, Chatti, Aono, Chakrabartty. "Intelligent Pavement Condition Assessment Using Piezo-Floating-Gate Sensors," 7WCSCM, 2018.

    [3] Hasni, Aono, Lajnef, Chakrabartty, Faridazar. "Toward Autonomous Self-Powered Self-Sensing Civil Infrastructure," NDE/NDT for SMT, 2018.

    [4] Hasni, Chatti, Lajnef, Chakrabartty, Aono. "Damage Progression Identification in Asphalt Concrete Pavements: A Smart Self-powered Sensing Approach," Advances in Materials and Pavement Prediction, Papers from AM3P, 2018.

    [5] Aono, Hasni, Pochettino, Lajnef, Chakrabartty. "Quasi-self-powered Infrastructural Internet of Things: The Mackinac Bridge Case Study," ACM GLSVLSI, 2018.

    [6] Kondapalli, Pochettino, Aono, Chakrabartty. "Hybrid-Powered Internet-of-Things for Infrastructure-to-Vehicle Communication," IEEE MWSCAS, 2018.

    [7] Zhou, Aono, Chakrabartty. "Gaussian Process Regression for Improving the Performance of Self-powered Time-of-Occurrence Sensors," IEEE MWSCAS, 2018.

    [8] Mehta, Zhou, Aono, Chakrabartty. "Self-powered Sensing and Time-Stamping of Tampering Events," IEEE MWSCAS, 2018.

    [9] Zhou, Aono, Chakrabartty. "Tamper Sensitive Authentication of Passive IoT Devices Using Self-powered CMOS Timers," SRC TECHCON, 2018.

    [10] Aono, Kondapalli, Lajnef, Pekcan, Faridazar, Chakrabartty. "Self-powered Sensors to Facilitate Infrastructural Internet-of-Things for Smart Structures," ANCRiSST, 2017.

    [11] Aono, Chakrabartty, Yamasaki. "Infrasonic Scene Fingerprinting for Authenticating Speaker Location," IEEE ICASSP, 2017.

    [12] Aono, Lajnef, Faridazar, Chakrabartty. "Infrastructural Health Monitoring Using Self-powered Internet-of-Things," IEEE ISCAS 2016.

    [13] Takeshita, Aono, Chakrabartty. "Low-Power Microcontrollers for RFID Tags," REU16, 2016.

    [14] Aono. "EMG-Based Speech Filtering & Classifier," Michigan Space Grant Consortium, 2015.

    [15] Aono. "Speaker Identification," Yamanaka Seminar, 2014.

    [16] Aono. "Acoustic Filtering for Signature Generation Applied to Geolocation," JSPS Postdoctoral Fellowship for Foreign Researchers Report, 2014.

    [17] Aono. "Extracting Speaker Features from Sub-Vocal Speech Using Jump Resonance Filtering," Michigan Space Grant Consortium, 2014.

    [18] Aono, Covassin, Chakrabartty. "Monitoring of Repeated Head Impacts Using Time-dilation Based Self-powerd Sensing," IEEE ISCAS, 2014.

    [19] Chakrabartty, Feng, Aono. "Gen-2 RFID Compatible, Zero Down-Time, Programmable Mechanical Strain-monitors and Mechanical Impact Detectors," SPIE, 2013.

    [20] Aono. "Jump Resonance Filtering for Biometrics via Electromyography in Sub-Vocal Speech," Michigan Space Grant Consortium Conference, 2013.

    [21] Feng, Aono, Chakrabartty. "Gen-2 RFID Compatible Energy Harvesting Sensor For Structural Health Monitoring Applications," Michigan State University Engineering Graduate Research Symposium, 2012.

    [22] Aono, Shaga, Chakrabartty. "Exploiting Jump-Resonance Hysteresis in Silicon Cochlea for Formant Trajectory Encoding," IEEE MWSCAS, 2012.

    [23] Aono. "Application Note: PCB Design with EAGLE," MSU Technical Report, 2011.

    Acknowledgments

    National Science Foundation GRFP/GROW (DGE-0802267 & DGE-1143954)

    University of Tokyo & Japan Society for the Promotion of Science (GR14001)

    Semiconductor Research Corporation (SRC)

    Metal Oxide Semiconductor Implementation Service (MOSIS)



    Michigan State University

    • PhD Committee: Shantanu Chakrabartty, Fathi Salem, Selin Aviyente, Wen Li, Richard J. Enbody
    • Collaborators: Nizar Lajnef, Wassim Borchani, Hassenne Hasni, Karim Chatti

    Washington University in St. Louis

    • PhD Committee: Shantanu Chakrabartty, Roger D. Chamberlain, Baranidharan Raman, William D. Richard, Xuan Zhang

    Adaptive Integrated Microsystems Laboratory