Necklace-to-Eye Magnetic Link Delivers Megabit Data to Smart Contact Lenses

Smart contact lenses have long hovered at the edge of science fiction: transparent displays floating in the field of vision, continuous glucose monitors woven into hydrogel, and sensors that track every flick of the eye. Yet one stubborn problem has kept these devices confined to laboratory prototypes. How do you stream megabits of data to a receiver that is only a few millimeters wide, sits on a blinking eyeball, and moves constantly relative to the rest of the body? A team of engineers at Purdue University now offers a strikingly simple answer: hang the transmitter around the wearer’s neck and let magnetic fields do the rest.

The research, published in Frontiers in Electronics by Sukriti Shaw, Mayukh Nath, Arunashish Datta, and Shreyas Sen, demonstrates a wireless communication link based on resonant magnetoquasistatic (MQS) coupling between a necklace-mounted transmitter coil and a tiny receiver coil embedded in the rim of a contact lens. Unlike conventional wireless power transfer, which demands that coils be nearly the same size and closely aligned, the Purdue scheme embraces a weakly coupled regime in which a 15-centimeter necklace coil talks to an 8-millimeter lens coil across roughly 15 centimeters of air and tissue. The result is a channel that tolerates head motion, blinking, and anatomical variation while sustaining data rates near 5 megabits per second.

The physics behind the approach hinges on a frequency window below 30 megahertz, where the team operates their link at approximately 26.8 megahertz. At this frequency, the largest dimension in the system, the 15-centimeter necklace, remains electrically small compared with the wavelength of roughly 11 meters. The electromagnetic fields stay confined to the reactive near field, radiation is negligible, and the coupling between the coils is almost purely magnetic. This magnetoquasistatic regime brings an unexpected bonus: because the relative magnetic permeability of ocular and facial tissues is essentially that of free space, the human body is largely magnetically transparent. The fields pass through muscle, cornea, aqueous humor, and eyelid skin with barely any perturbation, meaning the link budget depends on coil geometry and relative position rather than on the detailed composition of the wearer’s anatomy.

To quantify this behavior, the researchers built detailed three-dimensional finite-element models in Ansys HFSS, embedding the lens coil in a layered ocular phantom that included the cornea, aqueous humor, crystalline lens, and vitreous humor, with a skin sheet draped over the coil to mimic a closed eyelid. The simulations compared resonant operation in free space against operation with full head and torso tissue models. The received voltage was virtually identical in all cases, confirming that tissue absorption is negligible at these frequencies and that even the transient coverage of a blink does not measurably degrade the channel. The team notes that eddy currents in conductive tissue do introduce a small, finite loss, but the effect is modest and already captured in the complex permittivity models used in the simulations.

Resonance itself proved critical to the link budget. When the transmitter spiral was driven directly without tuning, the received signal sat barely above the front-end noise floor. Adding series capacitors to bring both coils into resonance at 26.8 megahertz lifted the received voltage by roughly 35 decibels, pushing the signal more than 30 decibels above the assumed noise floor. The loaded bandwidth of the tuned link was 250 kilohertz, but the researchers showed that deliberately lowering the coil quality factors widens the bandwidth to 1 megahertz at a cost of only a few decibels of additional loss. This trade-off between peak signal and bandwidth turned out to be favorable: according to the Shannon-Hartley theorem, the wider channel more than compensates for the reduced signal, raising the theoretical channel capacity from 1.3 megabits per second to 4.6 megabits per second, and approaching 5 megabits per second at 2 megahertz of bandwidth.

Robustness under misalignment is where the MQS approach truly distinguishes itself. The team swept the transmitter tilt angle from parallel to perpendicular relative to the ground, varied lateral offsets up to 9 centimeters, and changed axial separations across realistic necklace-to-eye distances. Across all these perturbations, the path-loss variation remained around 10 decibels, and the theoretical channel capacity never dropped below 4.5 megabits per second in the scenarios considered. For a link whose receiver rides on a blinking eye while the transmitter swings with chest movement, this level of tolerance is remarkable. The channel capacity even increased slightly as the coils’ quality factors were lowered and the bandwidth expanded, a counterintuitive outcome that the researchers attribute to the wider spectrum outweighing the reduced peak signal.

Safety considerations, always paramount for devices worn on the eye, appear favorable. The team estimated specific absorption rate under a deliberately pessimistic assumption that the entire 10-milliwatt transmitter power, a conservative design point for their analysis, is absorbed in a 10-gram tissue volume in the orbital region. Even then, the estimated SAR remains at least a factor of two below the 2-watt-per-kilogram localized limit specified by international ICNIRP and IEEE guidelines for public exposure. In practice, the absorbed power near the eye will be far smaller, because the link budget shows substantial attenuation between the coils and most of the transmitted energy is stored in the reactive near field or dissipated in the circuitry rather than in tissue.

To validate the simulation results experimentally, the researchers assembled a benchtop prototype using a 20-centimeter single-turn transmitter loop and a 1-centimeter four-turn receiver spiral, both tuned to resonate near 26 megahertz with discrete capacitors. The transmitter was tilted to mimic a natural hanging necklace angle, and the receiver was placed 18 centimeters above and 5 centimeters to the side, a deliberately challenging geometry. The measured channel loss at resonance was approximately 60 decibels, consistent with the finite-element predictions for comparable dimensions and misalignment. When pieces of animal meat, serving as a tissue phantom with dielectric properties similar to human soft tissue, were placed between the coils, the channel loss changed by only a few decibels across the frequency sweep, confirming the magnetic transparency of tissue-like material at these frequencies.

The implications extend beyond the contact lens itself. The team frames the necklace-to-eye MQS link as a communication backbone for a broader ecosystem of on-body devices, part of what they describe as the internet of bodies, where wearables share data continuously without awkward alignment constraints or excessive power density near sensitive tissue. Compared with prior smart contact lens demonstrations, which relied on coaxial coils separated by mere centimeters or on higher-frequency links with tighter regulatory constraints, the MQS approach extends the communication envelope to tens of centimeters with arbitrary misalignment while keeping the receiver coil small enough to fit on a hydrogel lens rim.

Significant engineering challenges remain before a PC-on-the-eye becomes reality. The receiver integrated circuit must be co-designed with the lens coil under stringent silicon-area and power constraints, balancing front-end noise performance against the sub-milliwatt budgets typical of on-eye electronics. Long-term stability in the ocular environment, including exposure to tear fluid, repeated blinking, and potential corrosion of thin-film conductors, will demand careful materials selection and encapsulation. Formal compliance testing with anatomically realistic phantoms will be required before any productization. But the Purdue team has established something foundational: the communication channel itself, long considered a bottleneck for smart contact lenses, is unlikely to be the limiting factor. With a necklace as the transmitter and magnetic fields as the messenger, the path to continuous augmented reality and health sensing on the eye looks considerably clearer.

Subject of Research: Efficient communication channel for smart contact lens with resonant magnetoquasistatic coupling

Article Title: Efficient communication channel for smart contact lens with resonant magnetoquasistatic coupling

Article References: Shaw, S., Nath, M., Datta, A., & Sen, S. (2026). Efficient communication channel for smart contact lens with resonant magnetoquasistatic coupling. Frontiers in Electronics, 6, Article 1704891. https://doi.org/10.3389/felec.2025.1704891

Image Credits: AI Generated

DOI: 10.3389/felec.2025.1704891

Keywords: Efficient, communication, channel, smart, contact, lens, resonant, magnetoquasistatic, coupling, scientific research