PaperlayerView exact revision on arXiv

Source-linked AI summary

A Two-Element Capsule Antenna for Near-Field Polarization MIMO in Wireless Capsule Endoscopy

Pyry Kiviharju, Lauri Vähä-Savo, Clemens Icheln, Kenichi Takizawa, Katsuyuki Haneda

arXiv:2610.03263v1eess.SP

TL;DR

WCE telemetry must support high-resolution, high-frame-rate images, while short links inside the body challenge far-field, rich-multipath assumptions. This paper proposes a conformal two-element antenna with orthogonal current distributions for polarization multiplexing and evaluates links through full-wave simulations and phantom measurements. Median channel capacity is 1.70 to 1.96 times the single-antenna counterpart, with the proposed design also leading the compared conformal antennas at 915 MHz.

  • Problem

    WCE image resolution and frame-rate demands challenge wireless data rates, while short in-body links may not meet the far-field, rich-multipath assumptions used in prior WCE MIMO studies.

    Source: §I · p. 1

  • Method

    The paper proposes a conformal two-element 915 MHz antenna with orthogonal current distributions and evaluates its near-field MIMO channel using full-wave simulations and phantom measurements.

    Source: §I · p. 1, §II · p. 2

  • Results

    Median channel capacity was 1.70 to 1.96 times the single-antenna counterpart in both environments, and the antenna yielded the highest capacity among existing conformal two-element WCE antennas at 915 MHz in the human body model.

    Source: §I · p. 2, §V · p. 11

  • Takeaways & Limitations

    The results support the proposed antenna's suitability for polarization multiplexing in WCE scenarios and show improved channel capacity compared with single-antenna counterparts.

    Source: D. Comparison between Simulations and… · p. 11

Abstract

from arXiv · show

Multiple-Input-Multiple-Output (MIMO) technology is considered to be a viable method for spectral efficiency and data rate enhancement in wireless communications. This article presents a two-element capsule antenna with orthogonal current distributions for polarization multiplexing in a wireless capsule endoscopy (WCE) application at 915 MHz. The proposed antenna was evaluated in a MIMO link where it served as the transmitter and an existing on-body antenna was the receiver. The performance was assessed in two distinct environments: an anatomically accurate human body model using full-wave (FW) simulations, and a homogeneous liquid tissue phantom, using both FW simulations and measurements. The former was used to evaluate the performance in a realistic environment and to demonstrate the superiority of the proposed capsule antenna against existing two-element antennas in the literature, while the latter aimed at verifying the antenna design by comparison of simulated and measured MIMO links. We studied channel capacity and eigenvalues for the comparisons. The MIMO link with the proposed antenna provided nearly doubled median channel capacity relative to a baseline single-input-single-output link, consistently in the simulations and measurements. Our results prove the suitability of the proposed antenna for polarization multiplexing and, in terms of channel capacity, establish it as the best conformal two-element capsule antenna design at 915 MHz.

I. INTRODUCTION

The paper addresses WCE telemetry demands in a near-field setting where common far-field, rich-multipath assumptions may not hold. It proposes and evaluates a two-element antenna for polarization multiplexing, reporting higher channel capacity than single-antenna links and existing conformal designs.

I. INTRODUCTION

  • WCE image resolution and frame-rate needs create data-rate challenges for links transmitting from inside the body.Source: §I · p. 1
  • Prior WCE MIMO studies generally assumed rich multipath and far-field propagation, assumptions the authors consider likely invalid for short-distance links inside the body.The paper instead characterizes the link as near-field and similar to underwater line-of-sight MIMO.Source: §I · p. 1
  • The paper proposes a conformal two-element antenna with orthogonal current distributions for polarization multiplexing and evaluates its channel using full-wave simulations and experimental measurements.The method accounts for near-field radiation, which leads to a channel dominated by direct electromagnetic coupling.Source: §I · pp. 1–2
  • Median channel capacity was 1.70 to 1.96 times the single-antenna counterpart in both environments, and the antenna had the highest capacity among existing conformal two-element designs at 915 MHz.Evaluation used FW simulations in a realistic 3D body model and simulations and measurements in a muscle tissue phantom.Source: §I · p. 2

II. TWO-ELEMENT CAPSULE ANTENNA DESIGN

The antenna is a conformal, two-element slot design on a two-layer PCB, developed to fit a capsule while meeting matching and isolation targets at 915 MHz. Its design combines slot miniaturization, microstrip feeding, and control of the parallel slot sections that dominate coupling.

II. TWO-ELEMENT CAPSULE ANTENNA DESIGN

  • The conformal antenna uses two mirror-image slot elements on a two-layer PCB bent into a cylinder, with a 1.3 mm gap between the substrate's short sides.The substrate is 0.25 mm thick RO3010 with 18 µm copper.Source: §II · p. 2

A. Antenna Operation Principle and Design Evolution

  • A quarter-wave slot was the starting design; widening its feed-side slot created an impedance step, and two 90° bends in the narrow section further reduced its size.The high-impedance section was made wider than the low-impedance section because slotline impedance increases with slot width.Source: A. Antenna Operation Principle and… · pp. 2–3
  • An L-shaped microstrip feed on the opposite PCB side provides a balun, while a short-circuiting via tunes the resonance to the band of interest.Source: A. Antenna Operation Principle and… · p. 3
  • Parallel slots coupled tightly (−5 to −10 dB), unlike collinear slots (−17 to −19 dB); coupling was attributed mostly to parallel sections and increased rapidly as their length h increased.Source: A. Antenna Operation Principle and… · p. 3
  • The optimized antenna meets |S11| < −10 dB and |S21| < −20 dB at 915 MHz for 1 MHz bandwidth in a homogeneous intestine tissue phantom.Source: A. Antenna Operation Principle and… · p. 4

B. Characteristics of the Antenna

The proposed antenna maintains input matching across the intended band in different GI-tract tissues and excites orthogonal current modes. Its simulated maximum SAR in the intestine model is well below the stated 10-g limit.

B. Characteristics of the Antenna

  • |S11| remained below −10 dB across the intended band in 50 mm-radius phantoms of GI-tract tissues, indicating robustness to tissue-induced resonance shifts.Fig. 5 plots reflection coefficient magnitude by tissue and marks the 915 MHz ISM band.Source: B. Characteristics of the Antenna · p. 4, Figure 5 · p. 4
  • The slots act as coupling elements that excite orthogonal central current modes, which lead to orthogonal near-field radiation and polarization multiplexing capability.Fig. 6 shows the current distributions for both elements at different phases.Source: B. Characteristics of the Antenna · p. 4, Figure 6 · p. 4
  • 0.004 W/kg was the maximum SAR in the intestine model, below the 2 W/kg limit averaged over 10 g of tissue.The calculation assumed uncorrelated sources and −10 dBm maximum input power split equally across the two antenna ports.Source: B. Characteristics of the Antenna · p. 4

A. Preliminaries

The preliminaries define a narrowband MIMO channel through its channel matrix and describe channel capacity and rank using its eigenvalues.

A. Preliminaries

  • The narrowband MIMO model relates the receive voltage vector to the channel matrix, transmit voltage vector, and receiver noise vector.The channel matrix comprises forward scattering parameters between each transmit and receive antenna.Source: A. Preliminaries · p. 5
  • Channel capacity is estimated from the channel matrix using bandwidth, transmit signal power, and receiver noise power.The receiver noise power is obtained from bandwidth, temperature, Boltzmann's constant, and receiver noise figure.Source: A. Preliminaries · p. 5
  • The channel rank counts eigenchannels whose SNR exceeds 6 dB; the channel is full-rank when that count equals min(Nr, Nt), and otherwise rank-deficient.Capacity is represented as the sum of eigenchannel capacities, each determined by the magnitude of an eigenvalue of HHH.Source: A. Preliminaries · p. 5

B. Simulation setup

The link simulations pair the proposed capsule antenna with a dual-polarized on-body receiver in a 3D human model. The setup samples capsule orientations and evaluates links at two internal locations.

B. Simulation setup

  • The proposed capsule antenna served as the transmitter and a dual-polarized double-bowtie antenna as the receiver, forming a 2 × 2 MIMO link in the “Nelly” 3D human model.The channel matrix was simulated with the FEM solver.Source: B. Simulation setup · p. 5
  • The two capsule transmitter locations represented shallow and deep positions, with link distances of 142 mm and 79 mm, respectively.Propagation crossed intestine, muscle, fat, and skin; both distances were below the stated 5λmat far-field boundary for electrically small antennas.Source: B. Simulation setup · p. 5
  • The capsule orientation was randomized across 200 channel realizations to estimate stable capacity and eigenvalue percentiles.Rotation angles were sampled from specified uniform ranges around the base orientation; mesh refinement continued until H-matrix magnitudes changed by less than 0.25 dB.Source: B. Simulation setup · p. 5
  • A −43 dBm transmit power was chosen to meet the 500 µV/m field regulation at 3 m, providing a larger SAR safety margin and reducing RF interference relative to −10 dBm.The stated comparison is with the IEEE 802.15.6 standard transmit power.Source: B. Simulation setup · pp. 5–6

C. Scattering parameters, Channel Capacity and Eigenvalues

The proposed link maintains matching and low coupling across the 915 MHz band, while its simulated MIMO capacity substantially exceeds the best contained single-antenna channel. Eigenvalue results show that rank deficiency is limited to some Txa links, whereas all Txb links are full-rank.

C. Scattering parameters, Channel Capacity and Eigenvalues

  • Across the 915 MHz ISM band, both link ends had reflection coefficients below −10 dB and coupling levels below −20 dB.Figure 8 plots the simulated reflection and coupling scattering parameters for the capsule transmit ports and on-body receive ports.Source: Figure 8 · p. 6, C. Scattering parameters, Channel… · p. 6
  • 11 Mbit/s at Txa and 19 Mbit/s at Txb were the simulated MIMO median capacities, 1.90 and 1.96 times their respective best single-antenna medians.The corresponding best single-antenna medians were 5.8 Mbit/s at Txa and 9.8 Mbit/s at Txb.Source: Figure 9 · p. 6, C. Scattering parameters, Channel… · p. 6
  • All Txb links were full-rank, while 7% of Txa channels were rank-deficient at the −43 dBm total transmit power.Median eigenvalues were −44 dB and −50 dB for Txa, and −32 dB and −37 dB for Txb; the figure marks an SNR = 6 dB rank-deficiency threshold.Source: Figure 10 · p. 6, C. Scattering parameters, Channel… · p. 6

D. Comparison with Other Two-Element Antennas

In simulations at 915 MHz, the proposed antenna outperformed the compared conformal two-element antennas across channel-capacity percentiles. The comparison attributes its advantage to efficiency and near-field pattern alignment, rather than greater multiplexing capability than every alternative.

D. Comparison with Other Two-Element Antennas

  • The benchmark used −43 dBm total transmit power and 200 simulated links with varying capsule orientations at two Tx locations.The antennas were reconstructed as accurately as possible, with matching circuits added for and because some reported dimensions or material parameters were incomplete.Source: D. Comparison with Other Two-Element… · pp. 6–7
  • The proposed antenna outperformed the other conformal two-element antennas across all channel-capacity percentile levels in the 915 MHz simulations.The comparison used the same total transmit power and bandwidth.Source: Figure 12 · p. 7, D. Comparison with Other Two-Element… · p. 7
  • The proposed antenna’s total efficiency was −38.7 dB, versus −41.3 dB for and −41.4 dB for, while and had comparable eigenvalue differences and multiplexing capability.Antenna had nearly identical element polarizations and entirely rank-deficient channels; the proposed antenna’s differing polarizations better matched the double-bowtie dipole receiver.Source: D. Comparison with Other Two-Element… · p. 7

IV. COMPARISON OF MIMO LINK SIMULATION AND MEASUREMENT IN HOMOGENEOUS TISSUE PHANTOM

The homogeneous liquid muscle-tissue phantom was used to compare simulated and measured MIMO-link characteristics because the heterogeneous human model makes experimental comparison difficult.

IV. COMPARISON OF MIMO LINK SIMULATION AND MEASUREMENT IN HOMOGENEOUS TISSUE PHANTOM

  • The phantom comparison was intended to verify the antenna design by comparing simulated and measured MIMO-channel characteristics.A heterogeneous “Nelly” human model was not used for this comparison because it makes comparison with experiments difficult.Source: §IV · p. 7

A. Homogeneous Liquid Phantom Model

  • The simulated homogeneous muscle-tissue phantom measured 340 mm × 280 mm × 260 mm and approximated the measurement setup.Its electrical properties are referenced to.Source: A. Homogeneous Liquid Phantom Model · p. 7, Figure 14 · p. 8
  • The phantom model used Tx locations Txc and Txd, at Tx–Rx distances of 140 mm and 50 mm, respectively, both shorter than the 5λ_mat far-field distance.Simulations introduced random x, y, and z displacements drawn from a zero-mean normal distribution with 3.5 mm standard deviation to represent capsule-position uncertainty.Source: A. Homogeneous Liquid Phantom Model · p. 8, Table V · p. 8

B. Capsule Antenna Prototype

The manufactured antenna prototype was modified for coaxial VNA measurements, conformed into a cylinder, and encapsulated for use in the liquid phantom.

B. Capsule Antenna Prototype

  • Coaxial VNA connections required a via between the PCB’s front and back metal layers, and increased dimension b from 15 to 19 mm.The front and back PCB sides became the outer and inner sides, respectively, after cylindrical conformation.Source: Figure 15 · p. 8, B. Capsule Antenna Prototype · p. 8, Table VI · p. 8
  • The prototype used a 0.25 mm Rogers RO3010 substrate with 18 µm copper and was shown before and after cylindrical conformation.The manufactured prototype was conformed and its measurement cables were soldered in place.Source: Figure 16 · p. 8, B. Capsule Antenna Prototype · p. 8
  • A roughly 0.3 mm-wall PLA shell insulated the conformed PCB from phantom liquid, while an internal polystyrene plug minimized the air gap.A support structure reinforced the coaxial cables against bending from the phantom liquid’s buoyant force.Source: B. Capsule Antenna Prototype · pp. 8–9, Figure 17 · p. 9

C. MIMO Link Measurement Setup and Procedure

The phantom study compared simulated and measured near-field MIMO links across capsule orientations. Measurements showed higher median capacity than the best single-antenna links, while eigenvalue differences were attributed to measurement and prototype discrepancies.

C. MIMO Link Measurement Setup and Procedure

  • Phantom-immersed Tx-to-Rx links were measured using a four-port VNA at 10 dBm Tx power; the prototype remained fixed at Txc or Txd during each measurement, irrespective of cable bend angle.Cable and support-structure constraints limited possible orientations; measured link counts were 110, 70, and 20 at cable bend angles of 0°, 60°, and 120°, respectively.Source: C. MIMO Link Measurement Setup and… · p. 9

D. Comparison between Simulations and Measurements

  • Across the 915 MHz ISM band, each antenna element had input reflection below −10 dB and inter-element coupling below −20 dB.Simulated and measured Tx reflection values differed, likely because of manufacturing tolerances and air gaps between the antenna PCB and encapsulation, but the matching level was maintained.Source: D. Comparison between Simulations and… · pp. 9–10, Figure 19 · p. 9
  • 7.3 Mbit/s and 22 Mbit/s were the measured MIMO median capacities at Txc and Txd, respectively, 1.71 and 1.90 times the best single-antenna medians.The corresponding single-antenna medians were 4.3 Mbit/s and 12 Mbit/s; this trend was consistent with body-model simulations.Source: Figure 20 · p. 10, D. Comparison between Simulations and… · p. 10
  • The measured first and second eigenvalue medians were −51 dB and −55 dB at Txc, and −28 dB and −32 dB at Txd.First eigenvalues agreed well between simulation and measurement, but second eigenvalues differed; the median second-eigenvalue differences were 2.3 dB at Txc and 0.7 dB at Txd.Source: D. Comparison between Simulations and… · p. 10, Figure 21 · p. 10
  • The second-eigenvalue deviation changed capacity by 9%, which the authors considered low enough to verify the antenna design and link evaluation methods.Cable shadowing not represented exactly in simulation contributed to some low measured second eigenvalues, particularly when cables lay partly between the Tx and Rx.Source: D. Comparison between Simulations and… · pp. 10–11

V. CONCLUSION

The paper concludes that its two-element capsule antenna supports polarization multiplexing in the WCE context. Simulated and measured phantom results agreed, and simulations showed higher capacity than single-antenna and existing conformal two-element designs.

V. CONCLUSION

  • The antenna improved channel capacity over the single-antenna counterpart by a factor of 1.70 to 1.96 in both evaluated environments and had the highest capacity among existing conformal two-element WCE antennas at 915 MHz in the body model.The 2 × 2 channel was full-rank in 97% of simulated body-model cases, averaged over deep and shallow Tx locations.Source: §V · p. 11
  • The conformal two-element antenna uses orthogonal current distributions on a shared aperture to support polarization multiplexing in the near field at 915 MHz.Its evaluation used an anatomically realistic human body model and a homogeneous muscle tissue phantom.Source: §V · p. 11
  • The authors identify distributed on-body arrays, additional capsule elements, and other medical-device bands such as 433 MHz as future directions.The stated suitability for WCE assumes channel state information is available at the transmitter.Source: §V · p. 11
Loading 2610.03263v1…