Search

This brief presents an E-band high-efficiency power amplifier (PA) implemented in a 28-nm bulk CMOS technology. The output matching network (OMN) is optimized for high power and high efficiency through a co-optimization technique where the optimum load admittance (Yopt) is derived by incorporating the loss of the OMN with the active device performance. In the mm-wave band, a transformer with a single turn ratio is widely used for OMN but poses a limitation on the maximum achievable conductance (Gout.max) to be seen by the output transistor. Therefore, a shunt capacitor (CL) is employed in this brief to boost ${\mathrm{ G}}_{\mathrm{ out.max}}$ up to the value required for reaching Yopt. Finally, the transformer dimension and ${\mathrm{ C}}_{\mathrm{ L}}$ are optimized for high efficiency and power through the co-optimization technique. The measurement shows that the PA exhibits more than 13.4 dBm of saturated power from 65 to 80 GHz. A peak power-added efficiency (PAE) of 27% is achieved at 71 GHz, which is the highest PAE among the reported E-band CMOS PAs to the best of authors’ knowledge. The PAE is maintained higher than 20% from 65 to 75 GHz. The PA exhibits a rms EVM of −25.8 dB, average PAE of 8.7%, and average output power of 5.8 dBm with a 3-Gbps 64-QAM signal.

Although studies have demonstrated the potential of using a single inertial sensor for walking and running motion prediction, further analysis of sensor position and training dataset size is required. This study investigates the accuracies and two types of uncertainties associated with using a single virtual accelerometer for whole-body motion predictions. Motion capture data during walking and running on a treadmill at four speeds were obtained from 198 participants. A Bayesian neural network (BNN) was trained to predict kinematic gait parameters from virtual accelerometer data. The root-mean-squared error (RMSE) ranged from 1.7° to 10.0° depending on three sensor positions– left wrist, lower back, and right shank. As the dataset size increased from 20 to 160 subjects, error and epistemic uncertainty decreased ( ${p} < 0.001$ ). However, aleatoric uncertainty remained constant ( ${p}$ = 0.247) based on the linear mixed-effect model. We observed sensor positions that predicted each gait parameter with the lowest error and uncertainty. For example, the sensor on the lower back showed less right hip flexion prediction errors (RMSE $4.04^{\circ }{)}$ than those on the left wrist (RMSE $5.10^{\circ }{)}$ and right shank (RMSE $4.29^{\circ }{)}$ . The two uncertainties corresponded to different sources of prediction error. The optimal sensor position was found for each target joint motion along with the uncertainties. We suggested a regression network that can quantify errors from two different sources, training dataset size, and system limitations such as sensor position. It can be used to improve data-driven motion prediction using wearable devices by understanding the limitations.

In this article, we analyze the mm-wave distributed amplifiers (DAs) with a modified artificial transmission line (M-ATL) model which includes the parasitic inductance and capacitance of the transistors. The characteristic impedance and the cutoff frequency are analytically derived with the M-ATL model. It is found that the parasitic inductance reduces the cutoff frequency, thus limiting the DA bandwidth and can be optimized for the group-delay equalization. To confirm the accuracy of the M-ATL model, two mm-wave DAs – ultra-wideband DA (UWDA) and variable-gain DA (VGDA) – were demonstrated using a 250-nm InP DHBT process. The on-chip bias network and group-delay equalization of the DAs are analyzed to improve the bandwidth and gain flatness. The measurement shows that the UWDA exhibits an average gain of 11 dB over the frequency from 6 to 261 GHz. The VGDA shows an average gain variation from 1.5 to 11.1 dB over the frequency from 6 to 200 GHz.

Hepatocyte growth factor (HGF) and its receptor, cMET, play critical roles in cell proliferation, angiogenesis and invasion in a wide variety of cancers. We therefore examined the anti-tumor activity of the humanized monoclonal anti-HGF antibody, YYB-101, in nude mice bearing human glioblastoma xenografts as a single agent or in combination with temozolomide. HGF neutralization, The extracellular signal-related kinases 1 and 2 (ERK1/2) phosphorylation, and HGF-induced scattering were assessed in HGF-expressing cell lines treated with YYB-101. To support clinical development, we also evaluated the preclinical pharmacokinetics and toxicokinetics in cynomolgus monkeys, and human and cynomolgus monkey tissue was stained with YYB-101 to test tissue cross-reactivity. We found that YYB-101 inhibited cMET activation in vitro and suppressed tumor growth in the orthotopic mouse model of human glioblastoma. Combination treatment with YYB-101 and temozolomide decreased tumor growth and increased overall survival compared with the effects of either agent alone. Five cancer-related genes (TMEM119, FST, RSPO3, ROS1 and NBL1) were overexpressed in YYB-101-treated mice that showed tumor regrowth. In the tissue cross-reactivity assay, critical cross-reactivity was not observed. The terminal elimination half-life was 21.7 days. Taken together, the in vitro and in vivo data demonstrated the anti-tumor efficacy of YYB-101, which appeared to be mediated by blocking the HGF/cMET interaction. The preclinical pharmacokinetics, toxicokinetics and tissue cross-reactivity data support the clinical development of YYB-101 for advanced cancer.