ZeroScatter-LoRa is an all-digital and zero-addedcomponent technique for Long Range (LoRa) backscatter communication using commodity Microcontroller Units (MCUs). Instead of using traditional Radio Frequency (RF) switches for impedance modulation, it utilizes existing digital peripheral pins of ultra-low-power MCUs, such as Inter-IC Sound (I2), Serial Peripheral Interface (SPI), or Inter-Integrated Circuit (I2), to produce LoRa chirp signals using logic-level changes. We show backward compatibility of ZeroScatter-LoRa with commercial LoRa Wide Area Network (LoRaWAN) gateways in multipathrich indoor environments with an average power consumption of $9.8 \mu \mathrm{A}$. The results show that the monostatic configuration using Universal Software Radio Peripheral (USRP) in full duplex mode has an achievable range of 2.5 m, and the bistatic configuration using a single-tone generator at a distance of 2 m (transmitterto-tag) has valid packet reception of more than 20 m (tagto-receiver). The technique is applicable to billions of existing MCUs using high-speed digital interfaces. The long range that is achieved using MCU pins for backscatter communication also suggests that security risks may be present in existing digital systems if digital I/O lines are not electromagnetically shielded from outside signals.

We present an automated real time approach for detecting and classifying vehicles in airborne synthetic aperture radar (SAR) images. This approach is based on a pre-trained You Only Look Once (YOLO) object classification model (YOLOv8n) that is re-trained using the Air Force Research Lab (AFRL) MSTAR dataset, which is a labeled corpus of X-band stripmap mode and spotlight mode SAR images having a typical range pixel resolution of 0.202 m and cross range pixel resolution of 0.203 m. The model is trained to detect 8 different vehicle categories against clutter consisting of both rural and urban scenes. The base model size is around 3.2 million parameters. The model was tested against a validation dataset consisting of an average of 279 instances of each vehicle class across 500 composite image frames consisting of clutter plus one or more vehicle target images. The model then provides each detected target’s bounding box center coordinates, bounding box height and width, predicted label, and confidence score unique to that image frame. The mean precision and recall scores across all 8 vehicle types is 0.75 and 0.79 respectively. The model runs at a frame rate of 31 frames/s on a 12-core 2.60 GHz Intel x86 processor with 32 GB of RAM.

We present a channel selective backscatter modulator for Bluetooth Low Energy (BLE) backscatter communication between wireless sensor devices and BLE host devices such as smartphones, tablets, or PCs. Conventional backscatter modulators have the potential to create spectrum pollution by modulating every signal incident over the entire antenna bandwidth. We demonstrate that adding a narrow-band bulk acoustic wave (BAW) bandpass filter (BPF) in line with the backscatter modulator suppresses unwanted out-of-channel backscatter modulation and thus reduces spectrum pollution. For example, adding a BAW BPF with a BLE Channel 39 passband reduces the unwanted backscatter of a BLE Channel 37 carrier by more than 69.7 dB, greatly reducing the potential for spectrum pollution. Finally, we validate this approach over-the-air in a multipath-rich environment, demonstrating that adding the channel selective BAW filter reduces the desired backscatter signal strength by only 0.78 dB on average while maintaining high opposite-channel suppression.

Integrated sensing and communication (ISAC) refers to the use of common hardware and software components to perform simultaneous sensing and communication tasks. In this work, a low-cost, low-power X-band (10 GHz) phased array radar system produces 2D synthetic aperture radar (SAR) images of a car, a motorcycle, and a person. Simultaneous communication alongside SAR imaging was demonstrated by adding a backscatter tag consisting of an RF switch and a WR90 horn antenna to integrate a binary phase-shift keying (BPSK) data uplink from the tag. The radar consists of an Analog Devices CN0566 phased array receive path, combined with an added WR102 standard gain horn antenna for the transmit path. The radar bandwidth is 500 MHz resulting in an SAR/ISAR image resolution of 0.3 m in range. The imaging was taken over $\approx 13^{\circ}$ in azimuth, yielding $\approx 0.065 \text{~ m}$ cross-range resolution. The resulting images were reasonable representations of the target size and shape, and the backscatter tag communication showed successful packet transfer with an expected range of up to $\text{1 2 ~ m}$ on boresight.

We present a two-channel frequency selective 2.4 GHz backscatter modulator for Wi-Fi (IEEE 802.11) and Bluetooth backscatter communication. Prior backscatter modulators create unwanted spectral pollution by backscattering every signal incident over the entire antenna bandwidth. In contrast, the proposed frequency selective modulator employs narrow-band bulk acoustic wave (BAW) filters having a 20 MHz bandwidth to enable backscatter modulation in either Wi-Fi channel 1 (2401-2423 MHz) or Wi-Fi channel 11 (2451-2473 MHz). The measured differential reflection coefficient |Delta Gamma| exceeds 0.568 across Wi-Fi channel 1 and 0.601 across Wi-Fi channel 11. The opposite-channel rejection exceeds 47.02 dB from Wi-Fi channel 1 to Wi-Fi channel 11 and 48.99 dB from Wi-Fi channel 11 to Wi-Fi channel 1. Both simulated and measured performance are presented and compared. When generating BPSK modulation at 11 Mbps (as used by IEEE 802.11b), the modulator has an average power consumption of 20.9 mu W and a measured energy figure of merit as low as 1.9 pJ/bit; meanwhile, when generating FSK modulation at 1 Mbps (as used by Bluetooth Low Energy), the modulator has an average power consumption of 1.9 mu W and a measured energy figure of merit as low as 1.9 pJ/bit. By addressing the key unsolved problem of backscatter-induced spectral pollution, this work paves the way for backscatter devices to become "first-class citizens" of the wireless spectrum.

We developed a low-power, X-band backscatter communication system using the ultra-low cost, commercially available HB100 Doppler radar module. We constructed two backscatter tags, each consisting of a bow-tie connected to a PIN diode, driven by an Arduino that implements frequency-shift keying (FSK) to backscatter a codeword to the reader. The HB100 radar module simultaneously acts as the continuous wave (CW) source, downconverter, and Doppler radar. This CW source frequency is set at 10.525 GHz with approximately 15 dBm output power. The reflected signal consists of the modulated backscatter signal as well as the Doppler signature of a moving target. The reflected signal is received by the HB100 and downconverted to baseband by its onboard mixer, amplified, and fed into a PC for real-time codeword identification and velocity tracking. We have verified a range exceeding 6 m in a multipath-rich environment.

We present an approach for achieving Bluetooth Low Energy (BLE) compatible wireless analog voltage telemetry using an all-digital approach. We use fully synthesizable digital logic based on Verilog register transfer language (RTL) code to implement all stages of the sensing pipeline, including analog-todigital conversion of a sensor voltage, data packetization in BLE data frames, and backscatter modulation to wirelessly uplink the sensor value to any existing BLE-compatible receiving device, such as billions of existing smartphones, tablets, and personal computers (PCs). The ZeroScatter backscatter communication approach leverages existing I/O pin drivers on FPGAs and standard-cell ASICs to implement wireless communication without requiring custom analog circuit blocks. In the ZeroScatter approach, backscatter modulation is achieved when a digital I/O pin is rapidly switched between input (high impedance) and logic-zero (low impedance) states, resulting in a time varying reflection coefficient at the terminals of the antenna. An initial demonstration using a TinyFPGA BX (Lattice ICE40LP8K) field programmable gate array (FPGA) consumes 1112 logic cells (LCs) out of the 7680 total (14.5% utilization) and has an estimated DC power consumption of 8.9 mW (total static + dynamic power consumption of the TinyFPGA BX) when sampling and telemetering an analog voltage between 0.9 V and 2.1 V and producing 8-bit samples at 137 samples per second. The measured communication range of this device exceeds 1 m to an iPhone XR when the ZeroScatter device is placed 1 m from a +15 dBm carrier source in the 2.4 GHz industrial, scientific, and medical (ISM) band.

Recent efforts toward achieving interoperability between backscatter communication devices and existing wireless communication standards such as IEEE 802.11 (WiFi) have driven the need for ever-wider bandwidth backscatter signals, up to 20 MHz or more for WiFi-6 for example. In this paper we consider a key problem in the generation of wideband orthogonal frequency division multiplexing (OFDM) backscatter signals as used in the IEEE 802.11ax standard. Backscatter modulation is achieved by presenting a carefully controlled reflection coefficient at the terminals of the device antenna. This reflection coefficient depends on the choice of backscatter modulator component values as well as the frequency-varying impedance of the backscatter device's antenna. For wideband backscatter systems, the frequency-varying antenna impedance thus presents a critical design tradeoff in the design of the backscatter modulator. We analyze the variation in backscatter signal power and the error vector magnitude (EVM) of backscatter signals across wideband channels given the measured antenna impedance variation of an example ceramic chip antenna in the 2.4 GHz industrial, scientific, and medical (ISM) frequency band. The analysis presented here is also applicable to narrowband signals that frequency-hop across a wide bandwidth, as is the case for Bluetooth and Bluetooth Low Energy (BLE) signals.

ZeroScatter BLE is an all-digital, fully synthesizable approach for adding Bluetooth Low Energy (BLE) backscatter wireless data uplinks to existing field-programmable gate arrays (FPGAs). Unlike conventional FPGA-based BLE devices that require an external BLE chipset in addition to the FPGA, ZeroScatter BLE is fully synthesizable from Verilog register transfer language (RTL) code. This allows ZeroScatter BLE to be be implemented on billions of existing FPGAs. A tri-state capable digital I/O pin, a stable digital clock that's divisible to 1 MHz, and a 2.4 GHz antenna connected directly to the FPGA I/O pin are required to form the BLE backscatter device. An external carrier source in the 2.4 GHz band is then backscatter modulated with a BLE subcarrier using the two distinct RF impedances formed by the FPGA I/O pin direction switching between logic-zero and tri-state. Our experimental setup includes a signal generator serving as a 2.4 GHz carrier wave (CW) source. This CW source is backscattered by an antenna attached to a Lattice iCE40 FPGA running the ZeroScatter BLE logic. The resulting BLE backscatter signal is then received by an unmodified iOS or Android device using a commercial BLE scanner app. With a CW carrier power of +30 dBm, we have verified a bistatic range exceeding 4 m in a multipath-rich indoor environment, using an unmodified iPad as the BLE receiver. This work points the way toward wireless communications links created entirely from commodity digital logic.

This study presents a demonstration of using a low-cost off-the-shelf X-band radar system for ISAR imaging of maritime vessels. The system consisted of the CN0566 kit from Analog Devices, a WR-102 horn antenna, and a computer. The radar receiver consisted of an Analog Devices CN0566 X-band dual 8 element linear array of 98 mm aperture, while the transmitter antenna was a 15 dB gain horn. The radar bandwidth was 500 MHz resulting in an ISAR image resolution of 0.3 m in range and 0.66 m in cross-range. Data was collected on two boats passing by the radar. Images were focused using simple track estimation, and the resulting images were reasonable representations of the vessel dimensions.

In this paper, we present a series of experiments with a 2.4 GHz dual-polarized electronic mode stirring system for mitigating the dense multipath observed in biomedical telemetry within metal animal cages, including automated control over the mode stirring configuration. Four dual-polarized mode stirring antennas establish eight bits of digital control over the mode structure in a 0.2 m metal cage volume. With an optimized mode stirring configuration, we observe 26 dB improvement in the worst case one-way path loss across the 2400-2483 MHz band at 2,124 surveyed locations, sampled with a 1 cm grid on the floor of the cage. Using an example Bluetooth Low Energy link budget, we compare three automated mode stirring strategies for re-configuring the mode structure in response to simulated Brownian animal motion within the cage. Without mode stirring, the link budget has margin in only 68% of the surveyed locations, while with mode stirring, the link budget has margin in 99% of the locations in a stationary-animal scenario, and 92% of the animal locations in a moving-animal scenario. Finally, we present a demonstration of the link margin improvement in an actual communication link using a backscatter-based Bluetooth Low Energy implementation.

This paper presents a frequency selective backscatter modulator for the 2.4 GHz industrial, scientific, and medical (ISM) band. Unlike conventional wideband backscatter modulators, which modulate every incoming signal that is incident on an antenna, a frequency selective modulator mitigates undesirable spectral pollution by preferentially backscattering signals over a desired range of frequencies and minimizing backscatter elsewhere. The modulator architecture presented here is designed for WiFi (IEEE 802.11) and Bluetooth backscatter devices operating in the 2400-2483 MHz ISM band. Three different band-selection filters are presented, based on microstrip and low-temperature co-fired ceramic (LTCC) implementations. Additionally, three different modulators based on varactor diodes, PIN diodes, and pHEMT FETs are presented. The wideband differential reflection coefficient is simulated and measured for each combination, demonstrating the frequency selective backscatter properties of the presented modulator architecture. For example, the pHEMT based frequency selective modulator exhibits a 5856 MHz 3 dB modulation bandwidth without filtering, while the addition of an LTCC filter reduces the 3 dB modulation bandwidth to only 328 MHz. This modulator has an energy-per-bit figure of merit of only 1.9 pJ/bit.

Delta-sigma modulation (DSM) enables the use of all-digital switched impedance modulators to generate arbitrary backscatter signals. For example, a DSM-based backscatter modulator having only two or four impedance states can generate quadrature amplitude modulation (QAM) with e.g. 64 states, or multicarrier orthogonal frequency division multiplexed (OFDM) signals having many subcarriers. This paper describes potential improvements to in-channel spectral characteristics by adding single sideband (SSB) and double sideband (DSB) noise shaping to the DSM signal path. Using numerical simulation and hardware validation, we demonstrate that noise-shaped DSM can improve the spurious-free dynamic range (SFDR) of OFDM subcarriers generated by a low-resolution two-state or four-state impedance digital-to-analog converter. The noise shaping approaches are validated using a prototype OFDM backscatter uplink based on an FPGA driving a single-pole-four-throw (SP4T) CMOS RF switch that serves as the backscatter modulator. The SSB and DSB noise shaping techniques are compared by over-the-air transmission of five-subcarrier OFDM backscatter symbols with a four-times oversampling DSM at up to 1.25 Mbps. With this approach, we find that DSB noise shaping yielded a 6.2 dB improvement in SFDR relative to SSB noise shaping, at the cost of 9.8 dB higher peak out-of-band quantization noise. These results confirm that an all-digital modulation approach with noise-shaped DSM can be used to balance in-band vs. out-of-band quantization noise and thus optimize the spectral characteristics of hardware-efficient, all-digital backscatter modulators for low-power wireless communication.

We demonstrated a nonvolatile electrically reconfigurable metasurface based on low-loss phase-change materials Sb 2 Se 3 with phase-only (~0.25π) modulation in the free-space. The tunable metasurface is robust against reversible switching over 1,000 times.

Free-space modulation of light is crucial for many applications, from light detection and ranging to virtual or augmented reality. Traditional means of modulating free-space light involves spatial light modulators based on liquid crystals and microelectromechanical systems, which are bulky, have large pixel areas ( 10 micron x 10 micron), and require high driving voltage. Recent progress in meta-optics has shown promise to circumvent some of the limitations. By integrating active materials with sub-wavelength pixels in a meta-optic, the power consumption can be dramatically reduced while achieving a faster speed. However, these reconfiguration methods are volatile and hence require constant application of control signals, leading to phase jitter and crosstalk. Additionally, to control a large number of pixels, it is essential to implement a memory within each pixel to have a tractable number of control signals. Here, we develop a device with nonvolatile, electrically programmable, phase-only modulation of free-space infrared radiation in transmission using the low-loss phase-change material (PCM) Sb2Se3. By coupling an ultra-thin PCM layer to a high quality (Q)-factor (Q 406) diatomic metasurface, we demonstrate a phase-only modulation of 0.25pi ( 0.2pi) in simulation (experiment), ten times larger than a bare PCM layer of the same thickness. The device shows excellent endurance over 1,000 switching cycles. We then advance the device geometry, to enable independent control of 17 meta-molecules, achieving ten deterministic resonance levels with a 2pi phase shift. By independently controlling the phase delay of pixels, we further show tunable far-field beam shaping. Our work paves the way to realizing non-volatile transmissive phase-only spatial light modulators.

It has previously been shown that the metal cages used for housing research animals resemble a reverberant cavity, presenting dense multi path interference for communication channels within the cage. This is due to the metal walls of such cages forming a resonant cavity having deep nulls at many locations within the cage volume. This creates significant challenges for neuroscience research where non-human primates are equipped with brain-computer interfaces (BCIs) to record neural activity. Prior work has shown that electronic mode stirring can be used to mitigate the deep nulls by selectively changing the electromagnetic boundary conditions on the cage walls. We present a novel dual-polarized 2.4 GHz electromagnetic mode stirring system consisting of integrated dual-polarized air dielectric patch antennas with CMOS RF switches enabling each polarization to be terminated in either a short or open condition and thus selectively changing the phase of reflection from the antennas. In initial testing over 72 surveyed locations within a test cage across the 2.4 GHz ISM band (2400–2483 MHz), dual-polarized mode stirring is shown to improve the worst-case two-way insertion loss between the BCI antenna and a cage-mounted antenna by 28.6 dB with dual-polarized mode stirring enabled, relative to without mode stirring. This approach is also shown to reduce the standard deviation of two-way insertion loss from 8.6 dB without mode stirring to 7.3 dB with dual-polarized mode stirring.

This paper describes an all-digital backscatter modulation approach leveraging delta-sigma modulation (DSM) to improve the in-channel spectral characteristics of orthogonal frequency division multiplexed (OFDM) backscatter communication. We demonstrate through numerical simulations and experimental validation that DSM can improve the spurious-free dynamic range (SFDR) of OFDM subcarriers generated by a low-resolution impedance digital-to-analog converter (DAC), such as an RF switch having two or four different impedance states. We present the design and validation of a prototype OFDM backscatter uplink with DSM implemented with all-digital logic in an FPGA. A single-pole-four-throw CMOS RF switch (i.e. 2-bits of impedance DAC resolution) serves as the backscatter modulator. We experimentally validated the DSM approach with a 2.4 GHz, five-subcarrier OFDM backscatter uplink and a four-times oversampling DSM at up to 1.25 Mbps. In this scenario, the DSM improved the SFDR by 43 dB within the subcarrier band while reducing the overall noise floor in the same band by 11.3 dB. These results confirm that a DSM approach can be used to control quantization noise and improve the spectral characteristics of low-resolution digital impedance modulators for backscatter communication in scenarios where in-channel SFDR is more important than wideband noise performance.

We present an all-digital application specific integrated circuit (ASIC) that implements Bluetooth Low Energy (BLE)-compatible backscatter communication. The ASIC was fabricated in a 65 nm CMOS process and occupies an active area of 0.12 mm2 while consuming a total of 205 $\mu \mathbf{W}$ DC power from 0.48 V and 1 V supplies. Of the total power consumption, 56% (115 $\mu\mathbf{W})$ ) is consumed by the digital logic, 16% (33 μW) by the on-chip clock oscillator, and 28% (57 $\mu\mathbf{W})$ by the RF switch used as a backscatter modulator. The ASIC broadcasts up to 1000 BLE advertising packets per second at a data rate of 1 Mbps, yielding a backscatter modulator efficiency of 57 pJ/bit, The device was validated in both cabled and wireless (over-the-air) measurement setups, demonstrating compatibility with unmodified smartphones as well as commercially available BLE chips, such as the Nordic Semiconductor nRF51822. With the wireless test setup used in this work, and assuming a +10 dBm carrier source, the ASIC has a theoretical maximum read range of 4.9 m using a smartphone as the receiver. Building from previous work in BLE backscatter communication using FPGA-based prototypes, this work provides an important quantitative demonstration of the size and power savings that can be achieved in a BLE-compatible backscatter ASIC.

We present a portable 3-D millimeter wave imaging system operating in the K-band (24 GHz). This imaging system consists of a multiple input, multiple output (MIMO) array of 32 transmit elements and 32 receive elements that illuminates a scene with millimeter wave energy and processes the reflected signals to reconstruct a 3-D image. To achieve an acceptable image resolution from this sparse array, the system combines multiple measurements while the sensor is moved relative to the scene being imaged. For ease of portability, the imaging system uses a single Ethernet cable to power the sensor and transfer the captured raw data to a laptop computer. A graphics processing unit (GPU)-optimized image reconstruction algorithm transforms the raw data to a 3-D image with approximately 1 cm voxel resolution, which is rendered in 3-D in a Web browser based user interface. We present measured test images and demonstrate an achieved dimensioning accuracy of ±1 − 3 mm when the system is used to detect and dimension objects hidden behind opaque building materials such as drywall, plywood, ceramic tile, vinyl flooring, and cement.

Neuroscience research in nonhuman primates (NHPs) often requires multiday neural recordings from freely moving animals inside their home cages, making ultralow-power uplinks using wireless backscatter communication highly desirable. Previous work reveals that the channel transfer function (CTF) of a standard NHP home cage in the 915 MHz and 2.4 GHz industrial, scientific, and medical (ISM) bands resembles a resonant cavity exhibiting deep nulls throughout the cage volume, which are particularly acute for round-trip backscatter paths. In this work, we investigate a novel application of passive antenna mode stirring via switched parasitic antennas (SPAs) to reduce the magnitude and prevalence of deep nulls in the cage CTF. We present a system leveraging four cage-ceiling-mounted SPAs with two dynamically controlled impedance states each, yielding 16 total mode stirring configurations. We compare the frequency-domain power ratio measurements at 126 positions throughout the cage taken with and without passive antenna mode stirring. In the 915 MHz ISM band, the optimized SPA configuration improved the maximum two-way insertion loss in 68% of testing locations, reducing the worst case two-way insertion loss by 60.2 dB. In the 2.4 GHz ISM band, the maximum two-way insertion loss was improved in 53% of testing locations, reducing the worst case two-way insertion loss by 35.6 dB. This approach eliminates the deepest nulls in the cage volume and leads to significantly improved link margin for a backscatter-based wireless brain–computer interface (BCI).