In this study, we investigated the device parameters of a gate-controlled carrier-injection silicon-on-insulator transistor (GCCI SOI-Tr) capable of achieving a steep subthreshold slope of less than 1 mV/dec. Six device parameters: base impurity concentration (N_b), base length (L_b), channel impurity concentration (N_ch), gate length, gate width, and argon implantation were varied. GCCI SOI-Tr has a unique structure with an inherent p-n-p-n thyristor and a connection between a p-channel metal-oxide semiconductor field effect transistor (MOSFET) and a lateral bipolar junction transistor (BJT) using a dummy gate. Accordingly, the dependence of the GCCI SOI-Tr characteristics on device parameters exhibited similarities to those of MOSFET and BJT. In particular, the amplification capability was increased by a lower N_b and shorter L_b, and the trigger voltage was controlled by N_ch. This indicates that the optimization of the GCCI SOI-Tr is possible by separating the optimization of the MOSFET and BJT components using a method similar to that of the conventional transistor.

This study investigates amorphous gallium oxide (a-GaO_x) thin-film UV-C photodetectors fabricated on synthetic amorphous quartz glass substrates using fine channel mist chemical vapor deposition (CVD) at 260°C. Post-deposition annealing (PDA) was performed at temperatures up to 800°C. X-ray diffraction confirmed that the films maintained their amorphous structure even after high-temperature PDA, in contrast to sapphire substrates, where crystallization and Al atom diffusion are typically observed. Optical measurements revealed that the optical bandgap (E_g,^opt) remained nearly constant after PDA, indicating that the quartz substrates effectively suppressed bandgap widening. PDA significantly reduced the dark current by more than two orders of magnitude, enhanced the UV-C wavelength selectivity, and suppressed persistent photoconductivity by shortening the response time. Device performance metrics, including responsivity (R), detectivity (D*), external quantum efficiency (EQE), and photo-to-dark current ratio (PDCR), were markedly improved, with the best results obtained for films treated at 800°C. These findings highlight the efficacy of combining quartz substrates with PDA to realize high-performance a-GaO_x UV-C photodetectors.

ReRAM-based Compute-in-Memory (CiM) architectures can accelerate AI workloads. However, device-level variations in ReRAM distort the distribution of multiply-accumulate (MAC) values, thereby degrading computation accuracy. This paper proposes a non-uniform quantization (NUQ) scheme optimized by genetic algorithms (GA) to reduce MAC readout errors while maintaining high computational efficiency. The proposed NUQ method adapts to resistance variations and signal aggregation effects. Under multi-weight and aging scenarios, it effectively mitigates signal overlap and improves readout accuracy by 60%.

Terrain-adaptive locomotion is an important issue for the autonomous walking robot to reach the destination within limited time and energy. To perceive the ground condition without image sensing and processing that consume a lot of energy, we investigated the artificial proprioception based on acceleration and the walking gait switching mechanism for an amoeba-inspired autonomous four-legged walking robot. We implemented the artificial proprioception mechanism by combining the 3-axis acceleration and a reservoir computing (RC) classifier. The RC model was trained offline using the data obtained from the robot walking on flat and rough grounds and the model was deployed on a commercially available microcontroller. We examined the ground condition classification in the four-legged walking robot with the trained model. We also demonstrated the switching of the walking gait in the robot between the pre-programmed normal walking and the amoeba-inspired successive searching for leg movement, depending on the ground condition. The gait switching based on our artificial proprioception was found to improve the locomotion efficiency of the robot walking on the complex ground conditions.

In this paper, a design methodology for wideband high-efficiency power amplifiers (PAs) is presented, employing a transformed radial stub (TRSs) filtering network integrated within the output matching network (OMN). This approach achieves simultaneous fundamental impedance matching and harmonic control, thereby significantly improving bandwidth and efficiency. Lumped-element equivalent circuit and odd-even mode analysis are employed to analyze the transformed radial stub (TRS). In addition, extended continuous Class-F mode (ECCF) is also employed to accomplish both wide bandwidth and high efficiency in PA design. The designed PA demonstrates wideband performance and high efficiency. To validate the proposed design strategy, a ECCF wideband and high-efficiency PA is fabricated and measured. The fabricated PA achieves a 137.5% fractional bandwidth (0.5-2.7 GHz), demonstrating a drain efficiency of 58.9-81.1%, an output power of 38.5-43.1 dBm, and a large-signal gain of 8.5-13.1 dB across the band.

This paper presents an adaptive integral sliding mode control (SMC) scheme for a multi-input multi-output (MIMO) hybrid energy storage system (HESS) which can be used in electric vehicles. In HESS, the battery acts as the main energy storage device, which is connected with a boost converter, and the supercapacitor acts as the auxiliary energy system, which is connected with a bi-directional DC-DC converter. The two converters share the same DC link for power exchange. Based on the two energy storage devices and connected converters, the global mathematic model of HESS is obtained. In the design process of the proposed control scheme, the adaptive estimation technique considering the unknown parameters of battery/supercapacitor hybrid system model is utilized. Moreover, with the combination of the projection algorithm, the parameters estimation function can be bounded. The Lyapunov based stability analysis of the whole battery/supercapacitor HESS is given to prove the convergence of the proposed adaptive integral SMC method. Finally, the effectiveness of the proposed method with a fast transient behavior is verified through the simulation test.

A Capacitance Varying Charge Pump (CVCP), a charge pump consisting of variable capacitors and rectifiers, has shown to generate an output voltage exponentially dependent on the stage number. Despite this high voltage generating capability, its output current is limited due to the small capacitance, when the variable capacitors are formed by Micro-Elector-Mechanical-Systems (MEMS) technologies. To address this issue, we propose two novel CVCP configurations that improve the current supplying capability. The first one employs electret-embedded capacitors that increase the charge capacity. The second one adopts a non-uniform capacitance configuration. It is shown that the effective charge pump resistance can be minimized when the capacitances are configured in an exponentially decaying manner with respect to the stage number. Theoretical analyses are validated by simulations, and the results demonstrate a significant improvement in output current.

Macro placement is a critical step in modern IC design, where reinforcement learning (RL) has shown promise by treating it as a sequential decision process. However, existing RL-based methods often suffer from low sample efficiency and insufficient exploration. To address these issues, we propose DSPLACE (Discrete Soft Actor-Critic PLACE), a macro placement approach that integrates Discrete Soft Actor-Critic (DSAC) with Efficient Channel Attention (ECA). DSAC enhances exploration through its maximum-entropy framework, while ECA strengthens state representation by adaptively recalibrating features. Experimental results on ISPD-2005 benchmarks show that DSPLACE improves wirelength by up to 14% and converges about 30.7% faster compared to recent RL-based counterparts.