PaperlayerView exact revision on arXiv

Source-linked AI summary

BioPhysical Modeling, Characterization and Optimization of Electro-Quasistatic Human Body Communication

Shovan Maity, Mingxuan He, Mayukh Nath, Debayan Das, Baibhab Chatterjee, Shreyas Sen

arXiv:1805.05200v1cs.ETcs.HC

TL;DR

Prior HBC studies report widely varying channel responses, motivating a model that accounts for measurement conditions. This paper develops a bio-physical circuit model and validates it with oscilloscope and wearable measurements. The results show nearly constant loss at low frequencies under the tested conditions and identify receiver termination as an important measurement choice.

  • Problem

    Prior studies report wide variation in measured HBC channel characteristics, and the effects of excitation and termination conditions need to be understood for channel characterization.

    Source: §I · p. 2, §II · p. 3

  • Method

    The paper develops a lumped bio-physical HBC circuit model and analyzes voltage-mode measurements using an oscilloscope and a miniaturized wearable prototype.

    Source: §I · p. 2

  • Results

    Wearable measurements show almost constant -49 dB to -52 dB loss across the frequency range, while oscilloscope results remain almost constant down to 10 kHz.

    Source: §IX · pp. 10–11

  • Takeaways & Limitations

    The model analysis identifies high receiver termination impedance and low source impedance as providing optimum channel loss for voltage-mode HBC.

    Source: §IX · p. 11

Abstract

from arXiv · show

Human Body Communication (HBC) has emerged as an alternative to radio wave communication for connecting low power, miniaturized wearable and implantable devices in, on and around the human body which uses the human body as the communication channel. Previous studies characterizing the human body channel has reported widely varying channel response much of which has been attributed to the variation in measurement setup. This calls for the development of a unifying bio physical model of HBC supported by in depth analysis and an understanding of the effect of excitation, termination modality on HBC measurements. This paper characterizes the human body channel up to 1MHz frequency to evaluate it as a medium for broadband communication. A lumped bio physical model of HBC is developed, supported by experimental validations that provides insight into some of the key discrepancies found in previous studies. Voltage loss measurements are carried out both with an oscilloscope and a miniaturized wearable prototype to capture the effects of non common ground. Results show that the channel loss is strongly dependent on the termination impedance at the receiver end, with up to 4dB variation in average loss for different termination in an oscilloscope and an additional 9 dB channel loss with wearable prototype compared to an oscilloscope measurement. The measured channel response with capacitive termination reduces low frequency loss and allows flat band transfer function down to 13 KHz, establishing the human body as a broadband communication channel. Analysis of the measured results and the simulation model shows that (1) high impedance (2) capacitive termination should be used at the receiver end for accurate voltage mode loss measurements of the HBC channel at low frequencies.

I. INTRODUCTION

Prior capacitive HBC measurements vary widely, in part because ground isolation and receiver termination alter measured loss. The paper develops a bio-physical model and wearable measurement approach to characterize the channel and evaluate signaling choices.

I. INTRODUCTION

  • Measurement setup can distort channel loss: insufficient ground isolation yields optimistic estimates, while low receiver impedance increases low-frequency loss.The paper identifies source and termination impedances as key factors in measured channel characteristics.Source: Figure 1 · p. 1, §I · p. 2
  • A lumped bio-physical capacitive HBC model explains channel loss up to 1 MHz across excitation and termination conditions and helps reconcile earlier measurements.The model is developed with experimental support and intended to explain discrepancies among prior studies.Source: §I · p. 2
  • The paper introduces a miniaturized, time-domain-sampling wearable prototype for voltage-mode measurements that approximates wearable-to-wearable communication.Source: §I · p. 2
  • Measurements from 10 kHz to 1 MHz characterize HBC as a broadband channel capable of supporting data rates up to 1 Mbps.Source: §I · p. 2
  • The study analyzes low- versus high-impedance and capacitive-versus-resistive termination, as well as low- versus high-impedance source choices, to recommend voltage-mode signaling conditions.Source: §I · p. 2

II. RELATED WORK

Prior capacitive HBC studies reported varied channel responses across frequency ranges and measurement setups. Their findings highlight the sensitivity of measured loss to ground connections, isolation methods, and electrode configurations.

II. RELATED WORK

  • Reported capacitive HBC responses range from bandpass to flat-band or low-pass, with substantial variation in loss across studies and setups.The paper identifies this setup sensitivity as a central issue for channel characterization.Source: §II · pp. 2–3
  • Cho et al. reported bandpass behavior from 100 kHz to 100 MHz, with loss reaching 90 dB at 100 kHz and passband loss varying with distance and ground-electrode size.Their grounded oscilloscope or spectrum analyzer provided a strong receiver ground connection.Source: §II · p. 2
  • Lucev et al. found that adding baluns changed a roughly 20 dB flat response into a bandpass response with about 80 dB loss near 100 kHz.Their network analyzer setup had common ground; baluns isolated transmitter and receiver grounds.Source: §II · p. 2
  • Other studies reported frequency- and setup-dependent responses, including bandpass behavior with balun isolation and a 6 dB average-loss change when receiver ground was connected.Hwang et al. observed the 6 dB change while retaining a flat-band response; Callejon et al. reported bandpass behavior with receiver-side balun isolation.Source: §II · pp. 2–3
  • The paper seeks to explain discrepancies in prior measurements with a unifying model and measurements up to 1 MHz for voltage-mode broadband HBC.The stated target is broadband communication supporting data rates up to 1 Mbps.Source: §II · p. 3

III. BIO-PHYSICAL HBC CIRCUIT MODEL

The bio-physical HBC circuit model represents electrode-skin contact, skin and internal body properties, and the receiver load. Its component values are informed by experiments and prior literature.

A. Human Body Forward Path Components

  • The model represents electrode-to-skin contact with a capacitance, while skin and internal tissue contribute resistive and capacitive elements.Tighter electrode contact increases the contact capacitance; the internal fat and conductive tissue impedance is on the order of hundreds of ohms.Source: A. Human Body Forward Path Components · pp. 3–4
  • The component inventory includes source resistance, body and ground-coupling capacitances, probe load capacitance, and load resistance.The listed load capacitance varies with receiver choice, including oscilloscope probes and the wearable receiver.Source: A. Human Body Forward Path Components · p. 4
  • Model component values draw on measured or literature-based properties, including electrode-skin contact and skin and tissue impedances.The model’s configuration also depends on excitation and termination modality; single-ended and differential excitation use different electrode and skin-impedance arrangements.Source: A. Human Body Forward Path Components · pp. 3–4

B. Return Path Capacitance Estimation experiment

The return-path capacitance experiment adds known capacitance between transmitter and receiver grounds and uses measured loss to estimate the unknown return-path capacitance. The estimated value is used in the HBC circuit simulations.

B. Return Path Capacitance Estimation experiment

  • Known capacitances are connected between transmitter and receiver grounds, and measured channel loss is used to estimate the return-path capacitance.The setup and equivalent circuit are shown in Figure 3; the loss relation uses the return-path and load capacitances.Source: B. Return Path Capacitance Estimation… · p. 4, Figure 3 · p. 4
  • The added experimental capacitance is combined with the return-path capacitances at the transmitter and receiver to represent the effective return path.Measurements across multiple known added capacitances allow the unknown return-path capacitance to be estimated.Source: B. Return Path Capacitance Estimation… · p. 4
  • The estimated return-path capacitance is around 1.5 pF, and this value is used in HBC circuit-model simulations.Source: B. Return Path Capacitance Estimation… · p. 5

C. Skin to Ground capacitance estimation experiment

The paper estimates body-to-ground capacitance as part of its HBC model and motivates low-frequency, voltage-mode characterization for broadband communication. Broadband HBC uses voltage signaling and time-domain sampling to characterize the channel toward 1 MHz.

C. Skin to Ground capacitance estimation experiment

  • The measured body-to-ground capacitance is around 150 pF, modeled as half at the transmitter and half at the receiver.The experiment estimates it from time-constant measurements with and without the body-to-ground capacitance in the circuit.Source: C. Skin to Ground capacitance… · p. 5

IV. BROADBAND HBC

  • A 1 MHz broadband channel can support data rates up to 1 Mbps, and broadband HBC avoids modulation and demodulation circuits.The paper notes potential energy efficiency of up to 6.3 pJ/bit.Source: §IV · p. 5
  • Broadband signals occupy bandwidth from DC to the data rate, unlike narrowband signals concentrated around a carrier frequency.This motivates characterizing the body channel at low frequencies rather than only near a carrier.Source: Figure 5 · p. 5
  • The paper uses voltage-mode signaling and time-domain sampling to characterize the channel for broadband HBC down to frequencies of tens of kilohertz.The received voltage is sampled to detect the signal, and measurements are compared with the biophysical model.Source: §IV · p. 5

V. COMMON GROUND MEASUREMENTS: HBC FORWARD PATH CHARACTERIZATION

Common-ground measurements isolate the human-body forward path from return-path effects and help identify model component values. The experiments sweep frequency and compare measured voltage loss with the biophysical model.

V. COMMON GROUND MEASUREMENTS: HBC FORWARD PATH CHARACTERIZATION

  • The common-ground experiment characterizes the forward path and compares measurements with the biophysical model to validate it.The section frames the measurements as a way to separate body-channel loss from return-path loss and determine circuit-element values.Source: §V · p. 5, A. Motivation · p. 6

A. Motivation

  • Connecting transmitter and receiver grounds removes the return-path capacitance effect so the measurement isolates the forward path.A signal generator supplies the input and an oscilloscope measures the received voltage.Source: Figure 6 · p. 5

B. Experiment Setup

  • The setup sweeps a sine-wave input across frequencies and measures voltage attenuation for single-ended and differential excitation.Voltage-mode measurements use a low-output-impedance transmitter and high-input-impedance receiver.Source: B. Experiment Setup · p. 6

C. Measurement Results

Forward-path loss varies with excitation and termination modality, while the measured results closely match the circuit model. The paper concludes that most HBC loss arises from non-common-ground return-path capacitance and that endpoint configurations affect loss.

C. Measurement Results

  • Approximately 0 dB loss occurs with single-ended excitation and termination, approximately 6 dB with differential excitation and single-ended termination, and approximately 10 dB with differential excitation and termination.Measured loss characteristics match the simulation model.Source: C. Measurement Results · p. 6, Figure 8 · p. 6

D. Conclusion: Channel Loss dominated by Return Path

  • Most HBC loss is attributed to non-common-ground return-path capacitance, while excitation and termination modality also directly affect forward-path loss.The paper says transmitter and receiver configuration and impedance should be chosen properly to minimize channel loss.Source: D. Conclusion: Channel Loss dominated… · p. 6

VI. NON COMMON GROUND: MEASUREMENT SETUP WITH CAPACITIVE RETURN PATH

  • The non-common-ground measurement section uses both an oscilloscope setup and a miniaturized wearable receiver to characterize channel loss and validate the model.The wearable setup is intended to represent HBC without measuring instruments that have a large ground.Source: §VI · p. 6

A. Oscilloscope based Setup: Large Ground

The oscilloscope setup provides high-impedance capacitive termination but its large ground can make measured loss optimistic. A miniaturized receiver uses ADC sampling, histogram-based amplitude estimation, and circuit de-embedding to measure a more wearable-like scenario.

B. Miniaturized Wearable Protoype based Setup

A. Oscilloscope based Setup: Large Ground

  • The 10x oscilloscope probe presents 10 MΩ in parallel with 13 pF, while the 1x probe presents 1 MΩ in parallel with 79 pF.Both provide high-impedance capacitive termination; the oscilloscope's large ground increases return-path capacitance relative to a wearable setup and can produce optimistic loss measurements.Source: Figure 10 · p. 7, A. Oscilloscope based Setup: Large… · p. 7

B. Miniaturized Wearable Protoype based Setup

  • Because the wearable transmitter and receiver lack a common ground reference, the received signal is biased before it is applied to the ADC.A resistive bias circuit provides this bias.Source: B. Miniaturized Wearable Protoype… · p. 8
  • The wearable receiver’s input load is modeled as a parallel resistance and capacitance.The circuit model represents the receiver input impedance.Source: Figure 13 · p. 8
  • The measured response is corrected for the amplifier gain and the bias circuit’s high-pass response to obtain the channel response.The amplifier gain is 12 over the measured frequency range, so the measured received amplitude is divided by 12.Source: B. Miniaturized Wearable Protoype… · p. 8

VII. MEASUREMENTS: BROADBAND HBC

The section compares oscilloscope and wearable-receiver measurements with the developed circuit model, using experiments across four male subjects over six days.

VII. MEASUREMENTS: BROADBAND HBC

  • Measurements covered 4 male subjects over 6 days.The transmitter was placed on the forearm and the receiver on the wrist.Source: §VII · p. 8
  • The forearm-to-wrist channel length was 20–25 cm.Source: §VII · p. 8
  • The section compares measurements from an oscilloscope and wearable receiver with the developed circuit model.Source: §VII · p. 8

A. Oscilloscope based Measurement Results

Oscilloscope measurements show flat-band response with capacitive receiver termination, while wearable measurements show greater loss but become flat-band after circuit de-embedding.

A. Oscilloscope based Measurement Results

B. Miniaturized Wearable Protoype based Measurement

  • Wearable measurements showed -49 to -52 dB channel loss, versus -43 dB and -47 dB with the oscilloscope’s 10x and 1x probes, respectively.De-embedding the amplifier and bias-circuit responses revealed a flat-band channel response; the higher wearable loss is attributed to lower return-path capacitance.Source: B. Miniaturized Wearable Protoype… · p. 9, Figure 15 · p. 9
  • The wearable prototype measurement includes a receiver amplifier and bias circuit whose frequency responses must be de-embedded to recover the channel response.The figure reports close agreement between the resulting response and simulation assuming 1pF input capacitance.Source: B. Miniaturized Wearable Protoype… · p. 9, Figure 15 · p. 9
  • Lower receiver termination impedance increases low-frequency loss and raises the channel cutoff frequency.The simulation comparison indicates that high termination impedance is needed to measure low-frequency channel characteristics.Source: Figure 16 · p. 9

A. Avoid 50Ω Termination

The section explains why low-impedance 50Ω measurement inputs overstate HBC loss, and how high-impedance capacitive termination reduces low-frequency loss.

A. Avoid 50Ω Termination

  • Previous VNA measurements showed about 20dB loss even without ground isolation, and the circuit model showed similar loss with 50Ω termination.The paper therefore advises against VNA or spectrum-analyzer measurements for voltage-mode channel loss.Source: A. Avoid 50Ω Termination · p. 9, Figure 17 · p. 9

B. Desired: High Impedance Termination

  • Because body impedance is a few kΩ, a 50Ω input causes high measured loss and a high-pass response whose cutoff rises as termination resistance falls.The paper recommends high-impedance measurement devices for more realistic voltage-loss measurements.Source: B. Desired: High Impedance Termination · p. 10

C. Capacitive Termination: Low frequency loss reduction

D. Voltage-Mode Signaling

Voltage-mode measurements favor high receiver impedance and low source impedance; the paper’s model and measurements support low-frequency broadband HBC characterization.

D. Voltage-Mode Signaling

  • Voltage-mode signaling favors high receiver impedance to maximize received voltage and low source impedance to reduce voltage drop and high-frequency loss.Conjugate matching is unnecessary at the discussed low frequencies because wavelength greatly exceeds channel length.Source: D. Voltage-Mode Signaling · p. 10

IX. CONCLUSION

  • The paper characterizes capacitive HBC up to 1MHz and reports nearly constant oscilloscope loss down to 10 KHz.Its lumped bio-physical model explains experimental loss characteristics for different excitation and termination conditions.Source: §IX · p. 10
  • 13pF and 79pF terminations produced -43dB and -47dB loss, respectively, while wearable measurements showed -49dB to -52dB across the frequency range.The conclusion identifies wearable-based measurements as necessary for accurate channel-loss estimation.Source: §IX · pp. 10–11
Loading 1805.05200v1…