A low-dropout linear regulator (LDO) serves as the core of power management, and automotive-grade applications require rapid transient responses across a wide range of load variations. However, conventional solutions face challenges in balancing a wide input voltage range with low power consumption and high load capability, exhibit slow transient response, and are prone to loop oscillation under light-load conditions due to an insufficient phase margin. This paper presents an LDO designed using a 180 nm bipolar-CMOS-DMOS (BCD) process, which integrates a load-transient monitoring compensation circuit with a dynamically driven Class-AB buffer. The design meets the requirement for fast transient response across a wide load range from 0 mA to 300 mA. At the input stage, a rail-to-rail constant-transconductance unity-gain buffer with a triple-current mirror was adopted. The buffer employs a dynamically biased Class-AB shunt feedback technique to enhance the slew rate at the gate of the power transistor. To suppress the overshoot caused by load variations, a discharge circuit was incorporated to limit the overshoot voltage to less than 4.8%. Additionally, the equivalent series resistance compensation network ensures a phase margin greater than 42° under a complex combination of three temperatures and three process corners, thereby guaranteeing loop stability.
As the special kind of one-dimension (1D) time series in terms of vector motions in the two-dimension (2D) plane (e.g., in oceanography), the bivariate signals play a critical role in various applications. This paper investigates bivariate signals via time-frequency analysis using uncertainty principles based on FRFT (the fractional Fourier transform). First, a generalized spectrum of bivariate signals is constructed using the newly proposed generalized QFRFT (the quaternion fractional Fourier transform), which differs from existing transforms. Then, generalized Heisenberg uncertainty principles in quaternion FRFT domains are derived, featuring tighter uncertainty bounds and more rational physical interpretations such as faster changing of bivariate trajectory reducing time-frequency resolution. Furthermore, windowed uncertainty principles based on the proposed windowed quaternion FRFT (WQFRFT) are explored for time-frequency analysis of linear frequency modulation (LFM) bivariate signals. Finally, the shown numerical experiments verify the effectiveness of the proposed relations.
Accurate field-programmable gate array (FPGA) power estimation is inherently time-consuming, often requiring iterative evaluations within power-oriented optimization loops, significantly prolonging the design turnaround time. Existing graph neural network (GNN)-based power estimation approaches predominantly rely on node-centric GNN architectures and necessitate register transfer level (RTL) implementation flows, which are computationally expensive and time-intensive. To address these limitations, we propose PowerMG, a novel dynamic power estimation framework that leverages early-stage information from electronic design automation tools, specifically from the high-level synthesis (HLS) front-end and back-end, bypassing the need for RTL-level data. We just need to pass through HLS flow to get the graph data and then feed it into the power estimation model, skipping the time-consuming RTL implementation stage in power inference. A key innovation of PowerMG lies in that it makes use of multi-grained GNN information. The power estimation value of each layer of our GNN is built on different GNN receptive field. Additionally, we introduce a complementary aggregation mechanism that integrates both edge-centric and node-centric layers, mitigating neighbourhood information loss and enhancing the representational capacity of the model. This dual-aggregation strategy is empirically validated to be effective for power estimation tasks. Experimental results demonstrate that PowerMG achieves superior estimation accuracy and strong generalization. It outperforms state-of-the-art methods on 8 out of 9 subsets of the PolyBench benchmark suite and surpasses PowerGear by 4.74% relatively at single model mode, highlighting its effectiveness and transferability.
Two-dimensional (2D) direction-of-arrival (DOA) estimation in single-snapshot super-resolution scenarios constitutes a pivotal challenge in radar detection for unmanned aerial vehicles. Existing algorithms can generally be categorized into two main types: Those based on atomic norm minimization (ANM) and those relying on enhanced matrix techniques. However, the ANM-based methods are constrained by the need for prior knowledge of noise power, while the enhanced matrix-based approaches suffer from a reduction in array aperture due to the reconstruction process of the enhanced matrix. To address these issues, we introduce a gridless version of the sparse iterative covariance-based estimation (SPICE) method, termed GLS, which estimates noise power in a vectorized atom form for 2D DOA estimation, leading to the development of a novel approach called V-GLS. Like ANM, V-GLS leverages semi-definite programming optimization, ensuring that the array aperture is not compromised. Additionally, we ingeniously reconfigure the 2D single-snapshot received signal vector into an approximated one-dimensional multiple-snapshot matrix, thereby decoupling the 2D DOAs into two separate one-dimensional DOAs. Moreover, we propose a decoupled version of GLS for 2D DOA estimation, referred to as D-GLS. Compared to V-GLS, D-GLS exhibits superior computational efficiency due to its low-dimensional semi-definite programming optimization. Both V-GLS and D-GLS exhibit robust performance in noisy environments and show promise for estimating multiple sources. Extensive simulations are provided to illustrate the superiority of the proposed method over the state-of-the-art methods.
With the development of intelligent driving and new energy automobile industries, automotive electronic systems are becoming increasingly highly integrated. As a result, vehicle-level power supply regulator chips will face more stringent load conditions. When dealing with complex load mutations, the output voltage is prone to overshoot and difficult to quickly restore stability, which seriously impacts the performance of lowdropout linear regulator (LDO). In turn, it causes problems such as delayed response of autonomous driving function and reduced stability of on-board electronic equipment. This paper proposes a transient enhancement circuit to improve LDO stability, significantly reducing overshoot magnitude and recovery time to normal voltage. Additionally, a current limiting circuit structure is introduced to effectively cap maximum load current and ensure circuit stability. Fabricated using the 180 nm bipolar complementary metal-oxide-semiconductor double-diffused metal-oxide-semiconductor automotive-grade process of United Nova Technology Co., Ltd., this LDO occupies an area of 0.530 mm2. The LDO supports a wide input voltage range spanning from 2 V to 5.5 V, delivers a maximum output current of 300 mA, and provides an adjustable output voltage range from 1 V to 3 V through configuration options. When the load current transitions abruptly from 300 mA to 0 mA, the overshoot voltage rapidly decays within 50 μs while maintaining a voltage drop of only 145 mV.
Differential analysis for block ciphers based on automated solving tools such as mixed-integer linear programming (MILP) and Boolean satisfiability problem (SAT), has become one of the key technologies in modern cryptographic research. Currently, existing automated analysis methods face efficiency bottlenecks when solving complex models, which limits their effectiveness in block cipher analysis. To address these challenges, we propose two improved modeling methods for differential analysis of block ciphers, aiming to enhance both solving efficiency and accuracy. Regarding the MILP approach, we introduce the “choice-based constraint” technique to model the propagation of differentials through linear layers. When applying this method to the 4-round differential analysis of uBlock-256, more compact and precise differential security bounds than existing results are obtained for the first time, demonstrating that 4-round uBlock-256 has at least 34 active differential S-boxes. For the SAT approach, we propose the “weighted encoding” method, which reduces the number of variables and constraints in the model, thus significantly improving solving efficiency. This method was employed in the optimal differential characteristic search process for 2 to 28 rounds of the PRESENT. Compared to the existing “sequential encoding” methods, the average solving time was reduced by 40%. The results have demonstrated the effectiveness and practicality of the proposed improved modeling methods in the automated differential analysis of block ciphers.
Integrated sensing and communication (ISAC) systems aim to unify sensing and communication within a shared waveform, offering benefits in spectrum and hardware efficiency. However, this integration raises critical security concerns, particularly regarding the leakage of sensing-related information due to the broadcast nature of wireless transmissions. While prior work has explored ISAC under data security constraints, securing sensing information remains a key open challenge. In this work, we address this gap by studying a two-receiver state-dependent bistatic ISAC model with a sensing eavesdropper. We characterize the fundamental trade-off among communication capacity, sensing distortion, and sensing information leakage. Specifically, we derive an inner bound using an input-constrained randomized encoding scheme and propose a new outer bound that incorporates leakage constraints. Under a reversely-degraded channel condition, we further establish the exact optimal trade-off. Numerical results validate our scheme and illustrate the complex interactions among capacity, distortion, and leakage in secure ISAC systems.
Aircraft trajectory prediction aims to forecast the future states of dynamic agents within a three-dimensional airspace, given their current and historical flight states. Intuitively, the historical flight states of these agents reveal the intentions of their pilots, which directly influence their future trajectory. However, existing approaches struggle to handle the long-term dependencies present in flight states and overlook the fundamental connection between an agent's behavioral intention and its trajectory. To address these challenges, this paper proposes a hybrid cached attention encoder-decoder (CAED) framework for aircraft trajectory prediction, with behavioral intention recognition as an auxiliary task. Specifically, CAED incorporates a dynamic cache within the encoder-decoder structure to capture the long-term dependencies between the targets' historical and current flight state representations. During the training process, an additive selection mechanism is proposed to select relevant history states stored in each memory slot that contribute most to the prediction of future trajectories. Finally, we design a scoring function and implement multi-task learning to jointly train our model. The extensive experiments on two aircraft trajectory prediction datasets clearly validate the superiority of our proposed CAED compared with various existing baselines.
Real-time detection of epileptic seizures can improve quality of life and support precision therapy. However, developing edge-based epilepsy monitoring systems remains challenging due to stringent demands for ultra-low power consumption, computational efficiency, and robustness. In this paper, a hardware-efficient, end-to-end nonlinear support vector machine (NLSVM) system for epileptic seizure detection is proposed. The system employs a frequency-time division multiplexed bandpass filter to extract key frequency-band features during epileptic seizures, followed by feature ranking and analysis using the minimum redundancy maximum relevance algorithm. By reformulating the Gaussian radial basis function (RBF) kernel with base-2 exponentiation and implementing a coordinate rotation digital computer (CORDIC) based co-design strategy for NLSVM hyperparameter optimization, the system achieves efficient RBF computation with reduced hardware overhead. The hardware-software co-designed feature selection module increases sensitivity by 2.7% through prioritized biomarker selection. Evaluated on the Bonn epilepsy database and Children's Hospital Boston-Massachusetts Institute of Technology Scalp epilepsy database using a Xilinx Zynq-7020 field programmable gate array, the system achieves sensitivity of 99.1% and 98.3%, respectively. Compared with Taylor-series and base-e CORDIC methods, the proposed classifier has less resource consumption, making it a hardware-efficient solution for real-time seizure detection.
In this work, a novel method for the theoretical derivation of general single nonuniform transmission lines is proposed by reasonably splitting complex propagation coefficient and characteristic impedance. The split quantities can at least be obtained as functions of line position by numerically solving corresponding Riccati ordinary differential equations. As a result, the voltage and current wave distributions along nonuniform transmission lines are derived in an explicit analytical form. Further on, the corresponding two-port network parameter expressions are formulated also in an analytical closed form. Both the voltage/current and network parameter expressions can be effectively simplified by adopting appropriate boundary conditions. Finally, the proposed method is applied to analyze a lossy linearly-tapered coaxial line and a lossy irregular coaxial line as two examples of nonuniform transmission lines. As a conclusion, the correctness, effectiveness, universality, and high efficiency of our theory and method are confirmed by the good agreement with the corresponding electromagnetic field simulation provided by the ANSYS high frequency structure simulator and the significantly shorter computational time compared to the conventional cascade method.
Designing antennas capable of simultaneously generating directive horizontally and vertically po-larized electromagnetic waves within a confined,radome-covered metallic cavity has been a long-term challenge.Traditional platform antennas are usually optimized in free-space environments,and when placed inside a metallic cavity,the cavity-induced reflections lead to degraded radiation performance.In this article,we propose a novel fully embedded structuralized-cavity antenna that achieves polarization diversity characteristics and forward-tilted radiation patterns.The proposed antenna consists of two orthogonally placed log-periodic dipole array radiators,a metamaterial-based absorber,a cylindrical cavity,and a metallic slope.Different from traditional platform an-tenna designs,this work considers the entire cavity structure,including internal configuration and radiator place-ment,as new degrees of freedom for radiation aperture design.A systematic design approach,involving radiator miniaturization,optimal radiator placement,and internal structure design of the cavity,is presented.Simulations show that the antenna exhibits forward-tilted radiation within the range of θ ∈[-80◦,-10◦]φ ∈[-45◦,45◦],,with a wide operating bandwidth from 2.9 GHz to 6.5 GHz for voltage standing wave ratio≤3.0.Measurements are in good agreement with simulations,demonstrating that the proposed antenna is well-suited for airborne applica-tions that require specific radiation patterns and stealth performances.
Crowdsourcing technology utilizes mobile sensing devices as basic sensing units to accomplish task allocation and data collection. Currently, it has emerged as a novel approach for addressing large-scale and intricate tasks. However, the collected sensing data often contain valuable information, necessitating careful consideration of data privacy. Traditional data privacy protection schemes are predominantly centralized, which makes them susceptible to single-point failures, trust crises, and scalability issues. Meanwhile, these traditional schemes fail to account for the risk of key leakage resulting from side-channel attacks within real network environments. Therefore, a distributed and leakage-resilient privacy protection mechanism has great practical value in crowd-sourcing applications. To this end, we propose a blockchain-based secure data sharing mechanism for crowdsourcing applications. Our main contribution is to propose an identity-based encryption scheme that achieves continuous leakage-resilient security. From the adversary's perspective, the ciphertext structure in the scheme seems completely random, making it infeasible to extract private key information. We establish the proposed scheme's chosen-ciphertext attack security under the selective-ID model, relying on the decisional weaker bilinear Diffie-Hellman inversion assumption. Theoretical analysis validates the scheme's advantages, feasibility, and effectiveness.
Even though the security of software architecture plays an important role in the security of complex and large-size software systems, there are still very few feasible methods to evaluate the security of software architecture quantitatively. To fill this gap, we propose an asset-oriented method to quantitatively evaluate the security of software architecture in this paper. The core idea is that we first use a multi-hierarchy dependence graph (MHDG) to describe the system and capture the dependence information between hierarchies, then we mark the positions of assets and attack points on MHDG, afterwards we identify potential attack paths and feasible attack paths from attack points to assets, and finally we evaluate the security of software architecture using some indicators from the perspective of software developers and defenders. We validated the method on four architectural styles and five evolutionary versions. Results show that it accurately captures the security differences among four representative architectural styles, effectively reflects security variations across five evolutionary versions of the same system, and facilitates security-oriented architectural optimization. Moreover, compared with random reinforcement strategies, our method achieves greater improvement in system security under the same protection cost, demonstrating practical value for security-aware architectural design.
Compared to traditional edge computing systems,edge-cloud hybrid systems are increasingly effec-tive in meeting the demands for low-latency and efficient task offloading.However,conventional edge-cloud com-puting often focuses solely on optimizing energy consumption,latency,or a combination of both,frequently over-looking the need for load balancing in task allocation among end devices,edge servers,and cloud servers.This pa-per explores an optimization framework that integrates the virtual machine(VM)architecture with an end-edge-cloud collaboration network(VM-ECN)for task offloading.We formulate a joint optimization problem that takes into account latency,energy consumption,and load balancing,facilitating task offloading through collaboration among end devices,edge servers,and cloud servers.To address it,we propose an improved multi-objective grey wolf optimization algorithm that incorporates tent mapping and a multi-strategy mechanism,referred to as tent mapping-based multi-objective grey wolf optimizer(TMOGWO).We conduct a comprehensive evaluation of the multi-objective optimization problem within the VM-ECN system using the analytic hierarchy process and per-form a series of experimental simulations with various parameters.The results show that our proposed TMOGWO achieves better convergence of composite indicator values across four distinct task scenarios compared to non-dominated sorting genetic algorithm Ⅱ(NSGAⅡ),multi-objective particle swarm optimization(MOPSO),multi-objective grey wolf optimizer(MOGWO).Furthermore,TMOGWO demonstrates superior performance compared to NSGAⅡ,MOPSO,and MOGWO in load-balancing metrics,while also achieving enhanced centralized distribu-tion and showing increased stability with respect to time delay and energy consumption.
This paper presents a broadband bandpass filter at Ka-band based on micro-coaxial structures. Micro-coaxial components get the advantages in small size, low insertion loss and high stability, which contribute to the integration and design of passive devices. The air-filled micro-coaxial transmission lines are used to design the proposed bandpass filter. It is composed of six paralleled resonant units and five quarter-wavelength transformation lines. Transition structures are designed to connect with other devices and facilitate the measurement. The designed filter is conducive to Ka-band communication system due to its compact structure and good performance. The center frequency of the designed filter is 28.4 GHz. The measured minimum insertion loss of the filter is as low as 0.601 dB and the 3-dB bandwidth is as broad as 56.9%. This indicates a good consistency with the simulation.
Inter-satellite handover management is critical to ensuring seamless service continuity and quality of service (QoS) in mega-scale satellite networks. However, traditional ground-centric handover architectures struggle with excessive signaling overhead and latency as networks scale to accommodate emerging services (e.g., remote control, broadband civil aviation, personal broadband). To address these differentiated services, we propose a space-ground distributed inter-satellite handover control architecture (SGDISHCA) that migrates radio resource control (RRC) and authentication management function (AMF) to satellites, establishing a distributed control paradigm. Specifically, we analyze the influence of the minimum elevation angle between the ground location area and the satellite on the delay and handover times, and establish the satellite-location area correlation model. According to the model, we design an on-demand allocation algorithm that dynamically aligns channel allocation rates with heterogeneous service demand rates, jointly optimizing QoS guarantees and resource efficiency. Through the analysis of simulation results, we validate that our proposed scheme reduces handover delays and signaling overhead. Furthermore, our scheme exhibits a remarkable enhancement in user satisfaction and optimization of system resource utilization. We also investigate the effect of ground station distribution on network performance and find that the network performance exhibits marked improvements with fewer and denser ground stations, facilitated by the on-demand allocation algorithm.
Cloud storage has transformed data sharing for individuals and organizations, but it also raises significant privacy concerns. Attribute-based encryption (ABE) enables fine-grained access control over encrypted data on untrusted servers, offering advantages over traditional symmetric or public-key encryption. However, with the advent of quantum computing, classical bilinear pairing-based ABE schemes face long-term security risks. Lattice-based ABE emerges as a promising post-quantum alternative, yet existing constructions suffer from large key/ciphertext sizes, inefficient Gaussian preimage sampling, and limited revocation support. To address these challenges, we propose trusted environment revocable ABE (TR-ABE), a lattice-based revocable ciphertext-policy ABE scheme explicitly designed for Intel Software Guard Extensions (SGX). By leveraging the SGX enclave, Gaussian SamplePre operations are securely outsourced while preserving confidentiality against side-channel leakage under the transparent trusted execution environment model. TR-ABE further adopts an attribute-aggregated key structure to enable efficient, resampling-free key updates, and supports efficient revocation with cryptographic binding of user identities to resist collusion. A cloud trusted execution environment framework ensures secure ciphertext updates with enclave integrity guarantees. We formally prove indistinguishability under chosen-plain-text attack security and collusion resistance under the ring learning with errors assumption in the standard model. Both theoretical analysis and experimental evaluation demonstrate significant improvements in revocation efficiency and up to 7.3× performance improvement in SamplePre through secure outsourcing in Intel SGX.
The emergence of non-volatile memory (NVM) has introduced new opportunities to enhance the performance of key-value storage systems. However, the high cost of NVM makes its exclusive adoption economically unfeasible. To address this issue, previous research has explored log-structured merge (LSM)-tree split architectures, aiming to achieve cost-effective performance improvements. Nevertheless, these studies often assign the lower levels of the LSM-tree to NVM and the higher levels to low-performance storage devices, which leads to significant compaction overheads at the higher levels and increased tail latency. In this paper, we propose an inter-leaving level placement scheme that maximizes device bandwidth utilization during compaction, thereby enhancing both read and write efficiency. Furthermore, we introduce a speculative read scheme designed specifically for this architecture, which reduces seek waiting overhead and improves random read throughput. Experimental results show that our scheme achieves up to 48.2% performance improvement over state-of-the-art systems.
State-of-the-art schemes for the performance analysis and optimization of multiple-input multiple-output (MIMO) communications generally suffer from degradation or even become ineffective in highly dynamic and complex environments with unknown interference and uncertain channel state information (CSI). To address these challenges and enhance network self-optimization, we propose a learnable model-driven regularized zero-forcing precoding scheme, and design a light-weight neural network for refined prediction of sum rate and detection error, by leveraging coarse model-driven approximations. Then, we estimate the CSI uncertainty based on the learned predictor in an iterative manner and, in turn, optimize both the transmit regularization term and subsequent receive power scaling factors. To achieve a favorable trade-off between convergence speed and robustness, we further propose a deep-unfolded projected gradient descent algorithm for power scaling.
