KOKenneth K. O.

Kenneth K. O.

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Kenneth K. O. has a total of 51 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • 400-GHz Concurrent Transceiver Pixels for Short Range FMCW Radar Imaging

    by Goutham Murugesan, Muhammad Awais, Sarfraz Shariff, Frank Zhang, Wooyeol Choi, Kenneth K. O.

    A 400-GHz concurrent transceiver imaging pixel with co-optimized output power, locking range (LR), and receiver noise performance is reported. The pixel integrates TX, RX, injection locking circuitry, and patch antenna within a ( $\lambda $ /2)2 area, which makes the pixel more easily scalable to form focal plane arrays. The transceiver pixel is fabricated in the GF 22-nm FDSOI CMOS process with a chip area of $740\times 520$ $\mu $ m and achieves a measured peak effective isotropic radiated power (EIRP) of −14.8 dBm, a measured minimum DSB NF of 33dB, and an LR of 8.5 GHz while consuming only 63.2 mW of DC power at 0.8-V VDD. The operation of the pixel with an FMCW injection signal is demonstrated over 8.4 GHz at a ramp rate of 174 MHz/ $\mu $ s. Using a module assembled with a PCB, including a pixel IC and a 6-cm-diameter Cassegrain reflector, radar imaging of a pair of corner reflectors at 1.85 and 2 m away with a cardboard cover is demonstrated. These design and characterization efforts were used to establish a design framework/flow for concurrent transceiver pixels.


  • Variation Reduction for Signal-to-Noise Ratios and Low-Frequency Noise of Far-Infrared Imaging Pixels Using an Array of Schottky Diode Detectors in CMOS and Postfabrication Selection

    by Muhammad Musab Farooq, Frank Zhang, Behnam Pouya, Kenneth K. O

    To mitigate the low frequency noise variability in Schottky barrier diodes (SBD’s) fabricated in a 110-nm foundry CMOS process and its effect on the performance of electromagnetic-thermal far-infrared (FIR) detectors, post-CMOS-fabrication selection of a subset of an array of parallel connected SBD detectors that form a pixel to reduce the noise and to increase the signal-to-noise ratio (SNR) is demonstrated. This technique at 20 THz reduces the SNR variation by ~4 dB compared to using all detectors in 5×5 arrays of FIR detectors with a 3.5 μm × 2.5 μm patch antenna. Furthermore, this technique is used to increase the worst-case SNR by ~5 dB. The sub-array combination of the 5×5 detector array with the highest SNR is utilized to create a thermal image of two 10-Ω carbon resistors with peak temperatures of 420 K and 380 K, respectively. The structural similarity index (SSIM) is employed to demonstrate that switching off varying subset combinations of detectors within a pixel does not degrade the image quality. The highest specific detectivity, D* of the pixels at 20 THz reported in this paper is 2.1 × 104, which is the highest reported D* for detectors not employing a thermal isolation cavity and operating at ~10 THz and above.

  • 305-GHz Cascode SiGe HBT Power Amplifier With L-C Feedback Achieving P sat of 8.3 dBm

    by Suprovo Ghosh, Haidong Guo, Frank Zhang, Kenneth K. O.

    A 305-GHz power amplifier (PA) fabricated in a 130-nm SiGe HBT BiCMOS technology with HBT $f_{t}/ f_{\text{max}}= 350/ 450\,\text{GHz}$ and Aluminum metallization is presented. The PA employs four-way combined four pseudo-differential cascode amplification stages with a capacitive feedback network between the collector of common base stage and the base of common emitter stage that counters the gain degradation by neutralizing the input loss and capacitance of common emitter stage resulting from the device parasitics, transit time related phase delay effects and the interconnect inductances in the layout especially due to the unavoidable interconnect between the output of common-emitter and input of the common-base stage. The PA achieves a measured $P_{\text{sat}}$ of 8.3 dBm, $OP_{\mathrm{1dB}}$ of 6 dBm, and a peak small signal gain of 14.5 dB at 305 GHz while consuming 880 mW of dc power from a 4-V supply. The PA exhibits the highest $P_{\text{sat}}$, $OP_{\mathrm{1dB}}$, and the highest small signal gain at 305 GHz among the PA's fabricated using SiGe HBT's with $f_{\text{max}}$ less than 500 GHz.


  • 10-GSymbols/s Supply-Modulated APSKC SiGe HBT Transmitter RF Front End for Operation at 250 GHz With 6.8-dBm Peak Modulated Power

    by Haidong Guo, Suprovo Ghosh, Frank Zhang, Suhwan Lee, Wooyeol Choi, Shenggang Dong, Ibukunoluwa Momson, Karl Strecker, Kenneth K. O

    This article presents a 250-GHz supply-modulated SiGe heterojunction bipolar transistor (HBT) ( $f_{T}$ and $f_{\max }$ of 350 and 450 GHz, respectively) transmitter RF front end that integrates a transformer-coupled four-stage power amplifier (PA) whose fourth stage is supply-modulated at a symbol rate of 10 GSymbols/s utilizing a high-speed low-dropout regulator (LDO). An integrated LO chain with a frequency multiplier generates a 121-GHz LO signal. Additionally, on-chip bipolar current-mode logic circuits are used to decode 1- or 2-bit PRBS data stream(s) and to drive the LDO that implements two- or four-level supply modulation. The transmitter supports 16-customized-APSK and 32-customized-APSK constellations, achieving instrumentation-limited data rates of 40 and 50 Gb/s with the peak modulated output power of 6.8 dBm and error vector magnitudes (EVMs) of −18.9 and −20.4 dB, respectively, while consuming approximately 1 W of dc power. The use of supply modulation with the APSK modulations increases both the peak modulated output power and RF-to-dc efficiency by $3{\times }$ and $2{\times }$ , respectively, compared to when conventional quadrature amplitude modulation (QAM) modulation with power backoff is employed. The measured 6.8-dBm peak modulated power is the highest for a transmitter IC operating at this frequency range.


  • 380-GHz Third Harmonic Signal Generation Using Differentially Pumped Varactors in a CMOS Voltage Controlled Oscillator

    by Zhe Chen, Hongcheng Dong, Zhiyu Chen, Wooyeol Choi, Ji-Xin Chen, Kenneth K. O, Wei Hong

    An LC-VCO incorporating differentially pumped varactors for 3rd order harmonic generation is demonstrated in a 65-nm bulk CMOS process with an fmax of ~250 GHz. The differentially pumped N-type Accumulation mode MOS varactor pair acting as a symmetric varactor serves as a frequency tuning element, as well as the non-linear reactive elements for harmonic generation. The measurements show that the circuit has a peak radiated power of -15.2 dBm with a frequency tuning range between 361.4 to 383.3 GHz or 5.9 %, and a peak DC-THz efficiency of 0.102 %. Among the CMOS oscillator signal generators with an on-chip antenna and output frequency greater than 300 GHz, the circuit exhibits the state-of-the-art DC-THz efficiency.


  • Post-Fabrication Performance and Longevity Improvement of a Phase-Locked Loop With an On-Chip Phase Noise Measurement Circuit

    by Farzaneh Jalalibidgoli, Pavan Yelleswarapu, Haidong Guo, Navneet Sharma, Joel Johnson, Tae-Yoon Kim, Frank Zhang, Behnam Pouya, K. O. Kenneth

    A 4-GHz phase-locked loop (PLL) incorporating an on-chip phase noise (PN) measurement circuit and a voltage-controlled oscillator (VCO) using an array of cross-coupled transistor pairs that can be configured using intelligent post-fabrication selection is reported. The cross-coupled pairs employ transistors with dimensions near the minimum for the process, and the post-fabrication configuration lowers its PN. The selection reduces the PLL PN at a 1-MHz offset from a 4-GHz carrier by $\sim$ 3.8 dB from the worst. The PN measurement circuit not requiring an external surface acoustic wave (SAW) filter was used to measure PN of $\sim$ $-$ 119 and $-$ 128 dBc/Hz at 1-and 2-MHz offsets, respectively. These are 2–3 dB lower compared to the lowest measured using on-chip PN measurement circuits in the literature. The maximum PN measurement circuit error at 1-MHz offset is 1.2 dB compared to signal source analyzer (E5052B) measurements. After high voltage stress that degrades PN, the performance was recovered by reselection, indicating that this design approach should also allow an increase in the VCO lifetime. This circuit is fabricated in a 65-nm CMOS technology with a chip area of 1 $\times$ 1.5 mm. The on-chip PN measurement circuit consumes 7.5 mW of dc power. This work suggests that designing for the best performance rather than the conventional approach of designing for the worst can be made practical.


  • Array of 296-GHz CMOS Concurrent Transceiver Pixels With Phase and Amplitude Extraction Circuits for Improving Reflection-Mode Imaging Resolution

    by Pranith Reddy Byreddy, Yukun Zhu, Suprovo Ghosh, Harshpreet Singh Bakshi, Walter Sosa Portillo, Jayson P. Van Marter, Kenneth K. O, Murat Torlak, Wooyeol Choi

    An 1x3 array of 296-GHz CMOS concurrent transceiver pixels incorporating an LNA, a differential gain stage, and a double-balanced I/Q mixer and a quadrature LO generator for extraction of baseband signal amplitude and phase in addition to the RF section in an area of ∼ (λ/2) 2 is demonstrated. The effective isotropic radiated power ( EIRP ) of the array is ∼-6 dBm and the minimum double sideband ( DSB ) receiver pixel noise figure is ∼48 dB. An E-shaped patch antenna is used to broaden the antenna bandwidth (14-GHz -10-dB |S11| bandwidth). Using a pair of these arrays, lens-less short-range reflection mode imaging of a target ∼1 cm away through a cardboard covering is demonstrated. The phase and amplitude outputs enable a synthetic increase of the effective aperture to the scan area from the pixel area for improved image quality and resolution. The use of phase combining for the 1x3 transmitter array also improves the isolation between the pair for reflection-mode imaging which limits the stand-off range by ∼10 dB (to ∼70 dB) compared to that when a pair of single pixels with a conductive wall is used.


  • Open Access

    Perspective on active submillimeter electromagnetic wave imaging using CMOS integrated circuits technologies

    by Kenneth K. O, Wooyeol Choi, Ruonan Han

    The performance of CMOS transmitters and receivers operating at the submillimeter electromagnetic wave frequencies have sufficiently improved for use in active transmission and reflection-mode imaging applications that have the potential for broad deployment and utilization. Imaging integrated circuits have the potential to be large in area to support a high number of pixels along with digital backend processing circuits. For high volume imaging applications that may eventually be included in automobiles, smartphones, laptops, tablets, and others, a large manufacturing capacity to support the volume of large area ICs is necessary. For this, the use of CMOS technologies with a much larger manufacturing capacity is favored. It should be possible to improve the performance of CMOS circuits to increase the range, and operation margin and frequency. The electronically steerable submillimeter-wave reflector technology holds the promise for improving the performance and energy efficiency of submillimeter-wave imaging systems by multiple orders of magnitude, and it is a critical research area. Increasing the operating frequency from 430 to 850 GHz using CMOS integrated circuits to improve the angular resolution by 2X at a given form factor (∼0.15° for a reflector diameter of 15 cm) can make the submillimeter-wave imaging competitive to the LIDAR angular-resolution performance, while providing superior capabilities in visually impaired conditions and making the imaging devices more affordable.


  • Open Access

    CMOS Platform for Everyday Applications Using Submillimeter Electromagnetic Waves

    by Kenneth K. O, Wooyeol Choi, Yukun Zhu, Haidong Guo

    CMOS (Complementary Metal Oxide Silicon) integrated circuits (ICs) technology is emerging as a means for realization of capable and affordable systems that operate at frequencies near 300GHz and higher. This is lowering a key barrier for utilizing the sub-millimeter electromagnetic waves in everyday applications. Despite the fact that the unity maximum available gain frequency, fmax of NMOS transistors (with connections to the tope metal layer) has peaked at 320 GHz, signal generation up to 1.33 THz, coherent detection up to 1.2 THz and incoherent detection up to 10 THz have been demonstrated using CMOS integrated circuits. Furthermore, highly integrated rotational spectroscopy transceivers operating at frequencies up to near 300 GHz, 300-GHz and 400-GHz concurrent transceiver pixels and arrays for high resolution radar imaging, and 300-GHz and 390-GHz transmitters, and 300-GHz receivers for high data-rate communication have been demonstrated in CMOS. The performances of these CMOS circuits are sufficient or close to being sufficient to support electronic smelling using rotational spectroscopy that can detect and quantify concentrations of a wide variety of gases; imaging that can enable operation in a wide range of visually impaired conditions; and high-bandwidth communication. Lastly, techniques for affordable packaging and testing sub-millimeter wave systems are suggested based on experimental demonstrations.


  • 426-GHz Concurrent Transceiver Pixel in 65-nm CMOS for Active Imaging

    by Yukun Zhu, Pranith Reddy Byreddy, Kenneth K. O, Wooyeol Choi

    This article demonstrates a 426-GHz concurrent transceiver imaging pixel in 65-nm CMOS that can be scaled to form a two-dimensional focal plane array. The pixel integrates a transmitter, a coherent receiver using a star-type triple-push oscillator, and an on -chip patch antenna into an area of 380×470 μm 2 . The pixel achieves a sensitivity of −89.6 dBm for a 1-kHz bandwidth with the third order subharmonic self-oscillating mixing in RX mode. This sensitivity is the lowest for imagers operating above 300 GHz. In TX mode, the pixel radiates −17.9 dBm. The pixel consumes 52 mW of dc power. If a 6-cm diameter Cassegrain reflector with a simulated gain of 47-dB gain is used, a link margin of 5.3 dB is expected for reflection-mode imaging of an object with a radar cross-section of ∼0.6 cm 2 at a range of 1 m.