The output voltages for the capacitive elements of a neural circuit model can be mapped into dimensionless capacitive variables, which present firing patterns similar to the membrane potentials detected in biological neurons. The inclusion of a memcapacitor also enables consideration of membrane deformation effects, enhancing the model's capacity to simulate neuronal behavior across varying physiological and environmental conditions. In this study, a capacitor and a memcapacitor are connected through a linear resistor in parallel with other electric components in different branch circuits composed of an inductor and a nonlinear resistor. The electrical activities in a neuron with a double-layer membrane and two capacitive variables are discussed in detail after converting the nonlinear equations for the neural circuit into a theoretical neuron model. A dimensionless neuron model and its corresponding energy function are derived. The field energy function for the neural circuit is converted into an equivalent Hamilton energy function and further validated via the Helmholtz theorem. Furthermore, the average value of energy serves as an indicator for predicting stochastic resonance, as supported by analyzing the distribution of the coefficient of variation. The neuronal firing patterns are shown to be energy-dependent. An adaptive control strategy is proposed to regulate mode transitions in electrical activities of the neuron. An analog equivalent circuit is constructed to experimentally verify the numerical results, thereby supporting the reliability of the proposed neuron model.

Vision Transformers (ViTs) have achieved remarkable success across various artificial intelligence-based computer vision applications. However, their demanding computational and memory requirements pose significant challenges for deployment on resource-constrained edge devices. Although post-training quantization (PTQ) provides a promising solution by reducing model precision with minimal calibration data, aggressive low-bit quantization typically leads to substantial performance degradation. To address this challenge, we present the truncated uniform-log2 quantizer and progressive bit-decline reconstruction method for vision Transformer quantization (TP-ViT). It is an innovative PTQ framework specifically designed for ViTs, featuring two key technical contributions: (1) truncated uniform-log2 quantizer, a novel quantization approach which effectively handles outlier values in post-Softmax activations, significantly reducing quantization errors; (2) bit-decline optimization strategy, which employs transition weights to gradually reduce bit precision while maintaining model performance under extreme quantization conditions. Comprehensive experiments on image classification, object detection, and instance segmentation tasks demonstrate TP-ViT's superior performance compared to state-of-the-art PTQ methods, particularly in challenging 3-bit quantization scenarios. Our framework achieves a notable 6.18 percentage points improvement in top-1 accuracy for ViT-small under 3-bit quantization. These results validate TP-ViT's robustness and general applicability, paving the way for more efficient deployment of ViT models in computer vision applications on edge hardware.

We propose an optimization method based on evolutionary computation for the design of broadband high-efficiency current-biased reverse load-modulation power amplifiers (CB-RLM PAs). First, given the reverse load-modulation characteristics of CB-RLM PAs, a comprehensive objective function is proposed that combines multi-state impedance trajectory constraints with in-band performance deviations. For the saturation and 6 dB power back-off (PBO) states, approximately optimal impedance regions on the Smith chart are derived using impedance constraint circles based on load-pull simulations. These regions are used together with in-band performance deviations (e.g., saturated efficiency, 6 dB PBO efficiency, and saturated output power) for matching network optimization and design. Second, a multi-objective evolutionary algorithm based on decomposition with adaptive weights, neighborhood, and global replacement is integrated with harmonic balance simulations to optimize design parameters and evaluate performance. Finally, to validate the proposed method, a broadband CB-RLM PA operating from 0.6 to 1.8 GHz is designed and fabricated. Measurement results show that the efficiencies at saturation, 6 dB PBO, and 8 dB PBO all exceed 43.6%, with saturated output power being maintained at 40.9–41.5 dBm, which confirms the feasibility and effectiveness of the proposed broadband high-efficiency CB-RLM PA optimization and design approach.

The rapid expansion of the low-altitude economy is driving strong demand for highly accurate and reliable positioning technologies to support diverse aerial operations. This review examines core positioning methodologies within the low-altitude intelligent network (LAIN) framework, beginning with an analysis of positioning requirements and performance metrics for low-altitude flight scenarios. It systematically assesses the principles, strengths, and limitations of mainstream positioning systems, including Global Navigation Satellite Systems (GNSS), terrestrial wireless positioning, and autonomous navigation, and it surveys prevalent integrated and cooperative positioning schemes. Our analysis demonstrates that standalone positioning technologies are inadequate in complex low-altitude settings, underscoring the pivotal role of multi-source fusion and unmanned aerial vehicle (UAV) swarm cooperative positioning as future trends. To address infrastructure gaps and high deployment costs in current LAIN systems, we propose a “space−air−ground” integrated and cooperative positioning architecture centered on GNSS and the 5^th generation mobile communication technology (5G). The ground layer integrates 5G and GNSS for wide-area enhanced positioning. The aerial layer uses 5G aircraft-to-everything (A2X) and sidelink (SL) communications to build self-organizing networks for cooperative UAV localization. The space layer leverages low Earth orbit (LEO) satellites to overcome coverage limitations in communication and positioning. This hierarchical architecture reduces deployment costs through infrastructure reuse and enables deep integration of communication and navigation capabilities. By supporting collaborative enhancement across all three domains, the framework improves positioning robustness and delivers cost-effective, ubiquitous, and highly reliable positioning services. Finally, we outline promising research directions. This review aims to provide a systematic reference and a novel architectural perspective for the ongoing development of LAIN.

As the frontier of multidimensional transportation systems, urban air mobility (UAM) is receiving increasing attention from international organizations, governments, and stakeholders in industry and academia owing to its high efficiency, low carbon footprint, and operational flexibility. Vertical take-off and landing (VTOL) infrastructure is the core facility that enables UAM and is therefore essential for its safe, efficient, and large-scale commercial implementation. However, the key technologies for establishing low-altitude VTOL infrastructure are still nascent, and government, industry, and academia have yet to harmonize the corresponding construction, management, and operation standards. To address this gap, we herein systematically review the related progress and trends, comprehensively surveying the key technologies of establishing VTOL infrastructure serving unmanned aerial vehicles (UAVs) and electric VTOL aircraft from three complementary perspectives of ground-side, airspace-side, and communication, navigation, surveillance, and information services. In the light of future UAM operations characterized by diverse vehicle types and dense air traffic, we propose a conceptual design for a public multioperator VTOL site to provide constructive insights into the sustainable growth of the low-altitude economy.