mmWave Full-Duplex Backscatter Tag Architecture

• mmWave full-duplex backscatter tag architectures are ultra-low-power devices that simultaneously receive a mmWave carrier and backscatter an uplink signal for integrated sensing and communications.
• They employ advanced RF components such as regenerative amplifiers, high-speed PIN-diode switches, and self-interference cancellation techniques to achieve multi-hundred-meter ranges and high data rates.
• Practical implementations balance antenna design, matching network complexity, and interference mitigation to deliver robust performance with a low bill-of-materials cost.

Millimeter-wave (mmWave) full-duplex backscatter tag architectures constitute a class of ultra-low-power, low-cost devices capable of simultaneously receiving a mmWave carrier (downlink) and backscattering an independently modulated, higher-frequency signal (uplink) toward a reader. Exploiting the abundant spectrum at 24–66 GHz and leveraging advanced RF front-ends—such as regenerative amplifiers, high-speed PIN-diode switch banks, and feed-forward self-interference suppression—these tags address the principal limitations of prior sub-GHz and single-duplex implementations by providing multi-hundred-meter ranges, high data rates, and rich signal processing capabilities for integrated sensing and communications (ISAC) scenarios (Chen, 2023, Harisha et al., 26 Jan 2026, Sun et al., 1 Apr 2025).

1. Architectural Principles and Operating Modes

The architecture of a mmWave full-duplex backscatter tag is grounded in “modulation-based” backscatter, wherein a passive or semi-passive tag dynamically adjusts its input impedance via high-speed switches or varactors to modulate the incident carrier with amplitude (OOK), frequency (FSK), or phase (PSK). Unlike relay-based or metasurface reflectors, which passively steer incident beams without information encoding, modulation-based tags natively support simultaneous reception (for energy harvesting or downlink data) and uplink modulation, directly enabling full-duplex operation (Sun et al., 1 Apr 2025).

A canonical mmWave full-duplex backscatter tag contains the following sequential modules (Chen, 2023, Harisha et al., 26 Jan 2026):

• Antenna module: patch array (e.g., 2×2 or 4×4) for beamforming and gain (8–12 dBi typical)
• Matching network: multi-element L–C–C or C–L–C topology, Q-factor 10–30
• RF front-end: 3-port circulator or directional coupler for isolation (≥20 dB)
• Switching/modulation circuit: PIN-diodes or varactors supporting OOK/FSK/PSK, switching time 0.5–2 ns
• Self-interference cancellation module: analog/digital feed-forward subtracter, 40–60 dB suppression
• Power harvesting: Schottky-diode rectifier (30–50% RF-DC efficiency at –10 to 0 dBm)
• Control logic: ultralow-power MCU or digital ASIC

This structure enables concurrent downlink reception and backscattered uplink with strong isolation between forward and reverse paths.

2. Block-Level and Circuit-Level Implementation

Recent prototype demonstrations utilize distinct paths for downlink and uplink, realized with independent antenna arrays and tuned RF circuitry (Harisha et al., 26 Jan 2026). Key differentiators include:

• Regenerative rectifier (downlink front-end): High-Q common-source pHEMT amplifier with positive feedback drives a Schottky-diode rectifier for ASK demodulation down to –60 dBm sensitivity, exceeding passive-only rectifier performance by over 50 dB.
• Uplink modulator: Regenerative amplifier (e.g., CEL CE3520K3 pHEMT, 30 mW quiescent), modulated via gate bias toggling for FSK (Δf ≈ ±2 MHz), achieves 30 dB gain and efficient re-radiation.
• Antenna subsystem: Triple patch-array configuration (downlink receive, uplink transmit, injection), each with ~8 dBi gain and >20 dB isolation via Wilkinson dividers or spatial separation.
• Digital baseband (MCU): Decodes ASK, drives FSK modulation logic, orchestrates duplex timing or multi-tag access.

This approach enables measured downlink BER of $10^{-1}$ at 200 m (ASK, 20 kbps) and uplink BER of $10^{-2}$ at 45 m (FSK, 500 bps–20 kbps) with a total tag power budget ≈112 mW and BOM cost <$5 (Harisha et al., 26 Jan 2026).

3. Theoretical Models and Signal Path Analysis

The mmWave full-duplex backscatter channel is characterized by the backscattered path impulse response, which in a bi-static configuration is:

$H_\mathrm{BS}(f) = \sum_{i=1}^{N_\mathrm{paths}} \Gamma(f)\,\sqrt{G_t\,G_\mathrm{tag}\,G_r}\;\frac{\exp\!\bigl(-j\,2\pi f\,(d_{t,i}+d_{r,i})/c\bigr)}{(4\pi\,d_{t,i}\,d_{r,i})}$</p> <p>where$d_{t,i}$and$d_{r,i}$denote respective Tx–tag and tag–Rx distances,$G_t, G_\mathrm{tag}, G_r$are antenna gains, and$\Gamma(f)$$is the instantaneous, possibly time-varying, reflection coefficient. For two-state switching:</p> <p>$$\Gamma(f) = \frac{Z_L(f)-Z_a(f)}{Z_L(f)+Z_a(f)}$</p> <p>where$Z_a(f)$is antenna impedance and$Z_L(f)$is the switched load. For ideal switching,$10^{-2}$0, $10^{-2}$1 (deep detune) or $10^{-2}$2 (180° phase shift) (Sun et al., 1 Apr 2025, Chen, 2023).

S-parameter characterization of the tag presents input return loss $10^{-2}$3, insertion loss $10^{-2}$4, and isolation $10^{-2}$5, with physical isolation (≥20 dB) and further digital cancellation to ensure that self-interference (SI) at the receiver is kept below –60 dBm (Chen, 2023).

The backscattered power at the receiver, assuming monostatic (co-located) Tx/Rx, is:

$10^{-2}$6

with $10^{-2}$7 ≈ 1–3 dBi and $10^{-2}$8 the propagation distance (Chen, 2023).

4. Full-Duplex Isolation and Self-Interference Mitigation

Full-duplex operation at mmWave frequencies mandates stringent Tx–Rx isolation due to the high magnitude of SI, especially where incident power is not negligible. Hardware isolation is achieved via:

• 3-port circulators or directional couplers (20–25 dB isolation; ~1 dB insertion loss)
• Orthogonal polarizations or spatially separated subarrays
• High-Q filtering, with frequency-planning (e.g., Δf = 320 MHz between uplink and downlink) providing ≥20 dB spectral selectivity (Harisha et al., 26 Jan 2026, Sun et al., 1 Apr 2025)

Analog SI cancellation is realized by tapping the incident carrier, shifting phase and attenuation to synthesize a cancellation tone, and subtracting it prior to the low-noise amplifier (LNA), yielding 40–60 dB SI suppression. Digital adaptive filtering (e.g., LMS or FIR approaches with known pilots) can further suppress residual SI (Chen, 2023, Sun et al., 1 Apr 2025).

The post-cancellation SNR is:

$10^{-2}$9

with $H_\mathrm{BS}(f) = \sum_{i=1}^{N_\mathrm{paths}} \Gamma(f)\,\sqrt{G_t\,G_\mathrm{tag}\,G_r}\;\frac{\exp\!\bigl(-j\,2\pi f\,(d_{t,i}+d_{r,i})/c\bigr)}{(4\pi\,d_{t,i}\,d_{r,i})}$0 the total suppression factor, typically targeting ≥60 dB (Sun et al., 1 Apr 2025).

5. Key Design Trade-Offs and Practical Parameter Space

Several trade-offs underlie mmWave full-duplex backscatter tag performance (Chen, 2023):

• Antenna size vs. bandwidth: miniaturized patch arrays (<5×5 mm²) restrict bandwidth (1–2%); wider-band approaches require multi-resonator or stacked structures (triple-resonator topologies).
• Matching network complexity: higher-order matching reduces VSWR but introduces excess insertion loss (0.5–1.5 dB typical).
• Isolation vs. loss: Circulators/couplers offer higher isolation at the expense of increased loss—and vice versa.
• Digital logic power vs. data rate: faster switching and higher constellation orders drive up consumption (1–10 μW for multi-Mbps), whereas low-leakage state machines enable sub-μW for sub-10 Mbps.
• Energy harvesting vs. SNR: duty cycle of $H_\mathrm{BS}(f) = \sum_{i=1}^{N_\mathrm{paths}} \Gamma(f)\,\sqrt{G_t\,G_\mathrm{tag}\,G_r}\;\frac{\exp\!\bigl(-j\,2\pi f\,(d_{t,i}+d_{r,i})/c\bigr)}{(4\pi\,d_{t,i}\,d_{r,i})}$1 must balance harvested power for on-tag logic and maximizing uplink SNR.

Empirically, practical parameter ranges include frequency bands 24–66 GHz, antenna gain 0–10 dBi, matching network Q of 10–30, PIN-diode switching time 0.5–2 ns, and energy harvesting threshold of –15 to –10 dBm (30% conversion) (Chen, 2023, Harisha et al., 26 Jan 2026).

6. Performance Metrics, Prototypical Tags, and ISAC Applicability

Recent architectures achieve unique combinations of range, sensitivity, and power/cost efficiency (Harisha et al., 26 Jan 2026). One system demonstrates:

Subsystem	Technology	Metric
Antenna	4×U-slot patch array	8 dBi, 12% BW (23.5–27GHz)
Regenerative Amplifier	GaAs pHEMT+R–C feedback	30 dB gain @ 30 mW
Rectifier	Feedback-enhanced Schottky	–60 dBm sensitivity
Uplink Modulation	FSK (±2 MHz)	10⁻² BER@45 m (FSK, 20 kbps)
Downlink Modulation	ASK	10⁻¹ BER@200 m (20 kbps)
Full-Duplex Isolation	Antenna + Q-filter	>20 dB
BOM Cost	PCB + actives	$5/tag
Power Budget	Tag total	112 mW

Integrated Sensing and Communications (ISAC) is readily enabled, as the tag’s backscatter preserves FMCW sweep phase, permitting sub-20 cm ranging with GHz bandwidth. Array processing (MUSIC/FFT) supports multi-dimensional localization, and frequency-division multiple access (FDMA) in narrowband regimes supports multi-tag scenarios (Harisha et al., 26 Jan 2026).

7. Research Challenges and Perspectives

Despite recent advances, formidable challenges remain (Sun et al., 1 Apr 2025, Chen, 2023):

• On-chip, wideband SI cancellation: At mmWave, passive duplexers are bandwidth-limited or physically cumbersome.
• Integrated MMIC implementation: Migration from PCB to CMOS/RFIC necessitates high-linearity, high-Q mixers, LNAs, and fast switch banks under severe power constraints.
• Wideband impedance matching: Simultaneous matching for harvesting, receive, and backscatter states over 20–30% bandwidth is non-trivial.
• Synchronization and channel estimation: Decoupling downlink pilots from uplink self-backscatter requires robust algorithms, especially with asynchronous multi-tag deployments.
• Standard compatibility: Coexistence within 802.11ad/ay regimes, spectral mask requirements, and multi-tag management present open system-level issues.

Existing reference designs such as mmTag (Van Atta SPDT array, 1 Gbps at 24 GHz), BiScatter (leaky-wave, 100 kbps full-duplex at 0.8 m), and MilBack (leaky-wave at 28 GHz, 40 Mbps) serve as the implementation basis for future integrated, battery-free full-duplex mmWave platforms (Sun et al., 1 Apr 2025).

A plausible implication is that scalable, low-cost, long-range mmWave backscatter tags can enable dense IoT, localization, and sensing infrastructures provided RFIC-level implementations, wideband SI cancellation, and robust multi-tag protocols are realized.