One rectenna, two polarizations, power plus identity

What the work claims

Djidjekh, Bernyer and Takacs claim a new rectenna topology: a 2-bit backscattering rectenna in the 868 MHz ISM band, built from two cross-polarized antennas (vertical and horizontal), an RF fail-safe single-pole double-throw switch and a backscattering rectifier, that operates in three modes selected by two GPIO control lines (Table 1): energy harvesting, co-polarized backscatter modulation through the rectifier MOSFET gate, and cross-polarized backscatter modulation through the RF switch¹. Measured RF-to-DC power conversion efficiency reaches 29% at minus 5 dBm input for the rectifier alone and a peak of 25% at minus 5 dBm with the fail-safe switch inserted (Fig. 2)¹. Integrated into a battery-free Bluetooth Low Energy sensor, the rectenna powered a complete sense-compute-encrypt cycle and backscattered a 16-byte AES-128 encrypted payload, a 3 ms Manchester-encoded frame intercalated before each BLE advertising event, successfully received and decrypted at 2 meters in cross-polarization and approximately 1.5 meters in co-polarization in an indoor environment¹.

How it works

A rectenna is an antenna plus an RF rectifier that turns incoming electromagnetic power into DC. A backscattering modulator instead varies the antenna load to create a modulated reflection of the incoming wave, communicating without any transmitter of its own. This paper's contribution is to make one aperture do both, and to separate the two functions by polarization. In harvesting mode both GPIO lines sit at 0 V: the fail-safe switch rests in its default state, routes the vertical-polarized antenna to the rectifier, consumes no DC power itself, and adds about 0.4 dB of insertion loss (Table 1, mode i)¹. In cross-polarized backscatter mode, GPIO 2 toggles the RF switch between the vertical receive antenna and the horizontal transmit antenna, on-off keying the reflected wave; switch losses between incoming and outgoing wave are around 0.33 dB, and the switch draws 1.8 mA only during the high states of the OOK sequence (mode ii)¹. In co-polarized backscatter mode, GPIO 1 drives the gate of a MOSFET (a BSS123) inside the rectifier itself, deliberately mismatching the rectifier input to modulate the reflection in the original polarization (mode iii)¹.

The rectifier is deliberately modest: a half-wave design with two Skyworks SMS7630 Schottky diodes, impedance-matched to a 10 kilohm load, fabricated on FR4. The authors state plainly that comparable published rectifiers at similar frequencies and power levels reach 24% to 56% peak efficiency, generally using optimized voltage-doubler or multi-stage topologies, whereas this design trades peak efficiency for the backscatter function and was measured at a fixed, non-optimal load, with the power management unit's maximum power point tracking expected to recover some of the loss in the full system (Fig. 2 discussion)¹. The battery-free sensor around it is commercial-grade: an Epeas AEM3090 power management unit, an NXP QN9080 BLE SoC, a 220 microfarad storage capacitor and a Texas Instruments HDC2080 temperature and humidity sensor, cold-starting from an empty capacitor and cycling between 1.2 V and 3.8 V storage thresholds¹. The demonstration link was a 24 dBm signal generator feeding a +9.2 dBi patch antenna, with a spectrum analyzer and a 2 dBi monopole as the backscatter monitor, sensor at 2 meters (Fig. 3)¹. The measured energy overhead of the security function is the number worth keeping: 52 microjoules per cycle for rectifier backscatter, 93 microjoules for the RF-switch mode, against a sensor operating cycle that is otherwise entirely powered by the beam (Fig. 5)¹.

The strongest case for it

The strongest case is systems-level: this device closes a loop that beamed-power architectures always leave awkwardly open. A wireless power receiver that can also talk back over the same beam, with cryptographic identity, turns a passive rectenna into a network node: the power provider knows which asset it is feeding, the asset can report its state, and the identification channel is protocol-agnostic to whatever radio the asset also carries¹. The polarization separation is the clever part, borrowed from polarimetric radar: because the backscattered telemetry leaves in the orthogonal polarization, it is separated from the incoming power wave and from clutter, and the paper notes this reduces cross-jamming in real environments (Section II)¹. The fail-safe switch is quiet engineering judgment: with no supply voltage at all, the device defaults to harvesting, so a dead or booting node still charges rather than stalling in a modulation state. And the energy arithmetic is honest in the right direction: 93 microjoules for a 3 ms encrypted backscatter burst is a marginal cost against a full sense-transmit cycle, meaning the security channel is nearly free once the power beam exists¹. For a technology family whose hardest problem is often justification of the infrastructure, a rectenna that pays for its own telemetry out of the beam it receives is a genuinely elegant result.

Where a skeptic should push

Start with the efficiency claim's measurement boundary. The 29% and 25% figures are single-device RF-to-DC conversion at 868 MHz, measured at the rectifier output into a fixed 10 kilohm resistive load at minus 10 to minus 5 dBm input power, on a bench, with a second tuning pass against the real power management unit¹. They are not system efficiency: no wall-plug number exists, the transmit side is a signal generator, and the wireless link budget is assumed rather than measured as a chain. There is no end-to-end efficiency from transmitter DC to sensor DC, so none can be quoted. The comparison the authors volunteer is also the softening: optimized multi-stage rectifiers at similar power levels reach 24% to 56%, and this half-wave design sits at the bottom of that band by construction, sacrificing conversion for the modulation function¹.

Second, the range and security results are demonstration-scale. The 2 meter cross-polarized and 1.5 meter co-polarized ranges were achieved indoors with no dedicated signal processing at the monitor and a 24 dBm source through a +9.2 dBi antenna; the authors list radar cross section analysis, polarization isolation evaluation and range optimization as future work, so the clutter-rejection advantage of the orthogonal polarization is asserted by analogy to radar, not quantified here¹. The security claim is architectural: a 16-byte AES-128 block acts as a dynamic private identifier, and the paper argues this hardens the BLE sensor against replay, relay and eavesdropping, but no attack, threat model or key-management analysis appears in the paper¹. Treat the cryptography as transport, not as a demonstrated security property. Finally, everything here is consumer electronics on FR4 at room temperature; nothing about the topology has been characterized for the thermal, radiation or vacuum environment that any space application implies.

What a two-mode rectenna means for beamed power

The non-obvious implication is that the receiving end of a beamed-power architecture can carry its own accountability. Every serious space beaming concept, from space solar power to laser-fed lunar surface assets, eventually faces the same dull question: how does the power provider verify what the beam is actually hitting, and how does the receiving asset report its condition? This paper's answer, at meter scale, is a mechanism rather than a subsystem: a rectenna that reports an authenticated identity in the polarization orthogonal to the incoming power wave, for tens of microjoules, using no transmitter and no extra aperture (Table 1, Fig. 4)¹. Translated to a beamed-power context, a surface asset fed by microwave or radio-frequency power could confirm its identity and report received power back through the same link, giving the beaming side a client-side measurement it currently has to model or assume. The same polarization trick that separates telemetry from clutter indoors is, in free space, a natural duplex channel between a power wave and its receiver.

The threat is extrapolation by scale, and it is serious. Every performance number here belongs to a meter-scale indoor link at minus 5 dBm: 25 to 29% conversion, 2 meters of range, microjoule budgets. A space beaming receiver operates at power densities, apertures and link distances where rectenna topologies, thermal behavior and efficiency curves are entirely different, and where a half-wave Schottky design at the bottom of the 24% to 56% band has no standing¹. Anyone who cites this paper as evidence for space wireless power infrastructure has confused a mechanism demonstration with a performance demonstration.

The opportunity, grounded in the paper's own setup, is beam-health instrumentation. The authors' backscatter monitor, a spectrum analyzer on a horizontally polarized antenna watching the cross-polarized telemetry at 2 meters (Fig. 3b), is a stand-in for what a beaming transmitter could do natively: interrogate a rectenna asset and read back an authenticated, payload-carrying reflection without the asset expending its stored energy on a radio¹. For distributed surface assets, say sensor nodes scattered across a lunar landing site fed from a single beamed source, that turns the power beam into a power-and-telemetry bus, and the 52 to 93 microjoule cost per report (Fig. 5) is the right order of magnitude to make reporting cheaper than silence¹. What would need to happen for that to be real is exactly what the authors defer: RCS and polarization-isolation characterization at realistic power levels, plus an honest end-to-end link budget.

The bottom line

Calibrated reading: high confidence that the mechanism works as shown, because the evidence is direct bench measurement of a fabricated device, and the authors are commendably frank about its limits. Moderate confidence that the polarization-duplexed power-plus-telemetry idea matters for beamed power in space, because the systems argument is sound and the energy cost is trivial, but no space-relevant performance claim can be drawn from this paper, and none is made here. The single result that would most raise the stakes is boring and decisive: a characterized radar cross section and polarization isolation for the cross-polarized backscatter mode, at power levels and apertures relevant to a real beaming link. Until then this is a neat, honest, meter-scale proof of concept whose architectural lesson outlives its numbers.