Electronics Letters最新文献

In massive MIMO–OFDM systems, the accurate acquisition of channel state information is crucial for ensuring reliable data transmission. As the antenna array size and signal bandwidth increase, wireless channel often exhibits sparsity in the angular and delay domains. However, the random distribution of spatial angles poses a persistent challenge for convolutional neural networks to extract spatial features effectively for channel estimation. To address this issue, we proposes a novel channel estimation network termed the attentive residual autoencoder network. Leveraging an autoencoder architecture enhanced with attention mechanisms and residual connections, the proposed method effectively captures frequency-space correlation. Numerical results show that the proposed algorithm significantly outperforms existing channel estimation methods.

Abnormal power usage detection is vital for power system security and efficient energy use. Existing methods are hampered by low sampling rates, insufficient feature extraction, and limited accuracy, which impede cost effective large scale deployment and raise privacy concerns. This paper proposes a low frequency electrical characteristic based abnormal power usage detection scheme. First, an abnormal power usage detection model is constructed by coupling active power and power factor under low frequency measurement conditions of existing smart meters. Second, the scheme embraces non-intrusive load monitoring (NILM) to safeguard user privacy and markedly reduce computational burden. Validation based on smart meter data indicates that the proposed scheme offers a more efficient, simpler, and more cost-effective solution for deploying advanced NILM-based metering systems in residential settings.

With the explosive growth of Internet of Things (IoT) devices, broadband signal detection in low-signal-to-noise ratio (SNR) non-cooperative environments faces severe challenges. Although deep learning has been widely applied to broadband spectrum signal detection, existing research on low-SNR environments remains insufficient. In such scenarios, the time-frequency representations of signals are highly susceptible to noise interference, resulting in difficulties in feature extraction and high false alarm and miss detection rates for existing deep learning models. Meanwhile, due to their high computational complexity, most deep learning models fail to meet the real-time processing requirements for massive data. Therefore, achieving high precision and fast detection of broadband signals under low SNR has become an urgent problem that needs to be solved. This letter introduces two lightweight modules, namely LFSI-DH and EOMFE, and integrates them into the YOLOv8 framework. Experiments demonstrate that the framework reduces complexity and improves accuracy compared to the YOLOv8, offering a viable edge device spectrum monitoring solution. Comparative analysis further validates its superiority in detection accuracy.

Using wireless passive sensor nodes as anchors is a cost-effective way for localization system. However, the quality of ranging information is determined by the power of request signal from the active targets. Thus, localization accuracy is influenced by the location request and transmission power allocation schemes. In this letter, we provide the Fisher information matrix of multiple active target localization using passive anchors. Further, we propose a joint collaborative beamforming strategy to optimize the localization accuracy considering the power constraints and localization availability constraints. The proposed strategy employs a two-stage semidefinite programming to fully exploit the spatial correlation. Simulation results demonstrate that our proposed strategy outperforms other schemes in improving the localization performance.

The traditional fuzzy c-means (FCM) clustering method is widely used in image segmentation and other applications. However, it is prone to noise and leads to under-segmentation. In this work, we propose a deep superpixel prior fuzzy c-means (DSP-FCM) clustering method for image segmentation as a new framework for expanding the deep neural network prior into existing iterative optimisation beyond traditional clustering models. The key insight is that the DSP is generated from an existing segmentation-based neural network, which enables us to derive an explicit variational formulation for capturing DSP and traditional superpixel prior together. Furthermore, a weighted local entropy is designed to minimise intra-class dispersion and maximise inter-class entropy. The overall framework suppresses noise and outlier influence and demonstrates superior performance through extensive experiments (on natural and medical image datasets) compared with existing state-of-the-art methods.

This letter introduces a dielectric-load slot antenna, which maintains an omnidirectional radiation pattern even when the size of the ground plane varies. The antenna comprises four slots arranged circumferentially and loaded with a strategically designed dielectric structure (DS). This DS reshapes the electric field distribution of the slots, forming an equivalent electric dipole that radiates omnidirectionally. Importantly, the reconfigured field and the resulting equivalent dipole remain unaffected as the ground plane enlarges, enabling consistent performance across different ground plane sizes. A prototype was fabricated and tested for validation. Measurement results show an omnidirectional pattern in the θ = 90° plane, with a maximum gain in that plane and a cross-polarization level below −10 dB.

A 12-bit 200 MS/s current-steering DAC is designed for low-voltage, low-power​ applications. By incorporating a partially randomised DEM (PRDEM) and ‘always-on’ current sources, the design simultaneously suppresses static mismatch and dynamic glitches while operating under reduced supply voltage. Measurement results show a 65.65-dB SFDR, which validates the effectiveness of the proposed techniques in achieving high dynamic range within strict power constraints.

To address the miniaturization requirement for W-band frequency modulated continuous wave radar front-ends, a transceiver chip was fabricated using 0.1 µm GaAs pHEMT technology. The design achieves 76–81 GHz broadband matching while resolving the compatibility challenge between high isolation and low loss in W-band monolithic front-ends. The chip integrates a mixer, directional coupler, filter and other modules simultaneously. A passive single-balanced mixer employing reverse-parallel Schottky diodes and a three-coupled-line Marchand balun structure serves as the core component. Measured at radio frequency = 76–81 GHz and intermediate frequency = 1 MHz–1 GHz, the chip achieves the conversion loss < 15 dB, local oscillation (LO)–intermediate frequency (IF) isolation > 36 dB and radio frequency (RF)–IF isolation > 25 dB.

This paper presents a wide-range, fast-lock duty-cycle corrector (DCC) with a 5-bit successive-approximation register (SAR). An inverter-based bang-bang duty-cycle detector (BBDCD) is equalised before each comparison to suppress hysteresis, enabling deterministic decisions and a fixed 4-cycle per-bit schedule. The duty-cycle adjuster (DCA) uses a controller frequency code for range adjustment and applies delay equalisation to limit code-dependent delay during updates. A half-LSB post-bias then halves the SAR quantisation-error bound without extra cycles. Post-layout simulations in 28-nm CMOS show operation from 0.8 to 3.2 GHz over 38%–62% input duty with a 20-cycle lock, ≤1.0% maximum duty error, and 1.73 mW at 3.2 GHz.

This letter proposes reflectionless linearly and circularly polarised patch antennas based on loading complementary absorptive branches. By establishing the equivalent circuit model of a conventional patch antenna, the complementary absorptive branch could be designed to dissipate the out-of-band energy through the grounded resistor, thereby achieving the reflectionless performance. Based on this concept, reflectionless linearly and circularly polarised patch antennas are designed and fabricated. Both antennas operate at 3.5 GHz. Measured results indicate that the linearly polarised patch antenna has a wide reflectionless frequency range of about 105% (1.71–5.4 GHz), with a peak radiation gain of 8.25 dBi. The circularly polarised patch antenna has a reflectionless frequency range of 68% (2.07–4.44 GHz), with the axial ratio (AR) bandwidth of 4.8%, stable radiation of right-handed circular polarisation (RHCP), and a peak radiation gain of 8.02 dBic.