The kT/C cancellation technique can effectively reduce the DAC size for SAR ADCs and thus relax the burden for input drivers and ref-erence buffers. However, the prior kT/C cancellation technique suffers from a hard trade-off between the noise, amplifier bandwidth and linearity. In this work, an improved kT/C noise cancellation technique is proposed to break the trade-off. It uses presampling to hold the input signal unchanged during the noise cancelling phase, leading to sig-nificantly relaxed requirements on the bandwidth and linearity of the amplifier. The proposed technique is verified in a 14-bit 200MS/s SAR ADC with the DAC size of 128 fF. Simulation results show that this work achieves 75.6 dB SNDR and consumes 1.56 mW power.

Point cloud data is a large collection of high dimensional 3D points with 3D coordinates and attributes, which has been one of the mainstream representations for emerging 3D applications, such as virtual reality, autonomous vehicles and robotics. Due to the large-scale unstructured high-dimensional nature of point clouds, point cloud processing, transmitting and analysing has been challenging issues in multimedia signal processing and communication. Deep learning is a powerful tool to learn statistical knowledge from massive data. Advances in artificial intelligence, especially deep learning models are offering new opportunities for point cloud processing, compression and analysis. This special issue aims at promoting cutting-edge research on deep learning-based point cloud processing, including object detection, segmentation, registration, compression, and visual quality assessment.

Silent Speech Recognition (SSR) based on Surface Electromyography (sEMG) is a voice interaction technology proposed for scenarios requiring silent operations. In this article, we abstract the SSR task based on sEMG into a short-term image sequence classification task. We perform time-frequency domain feature extraction and data reconstruction on the muscle activity segment data. Additionally, we analyze the temporal and spatial dimensions to capture the intrinsic correlation representation of muscle activity. We propose the SVIT-SSR model based on the Vision Transformer (VIT) framework. Finally, we design experiments to identify 33 types of typical silent speech commands in the SSR dataset. The results demonstrate that the proposed model achieves an accuracy of 96.67±1.15%, outperforming similar algorithms.

Retinopathy of Prematurity (ROP) mainly occurs in premature infants with low birth weight, and it is the leading cause of childhood blindness. Early and accurate ROP diagnosis is imperative for appropriate treatment. However, less research concentrates on early-stage ROP diagnosis based on limited-labeled images in an imbalanced dataset. To address the dilemma, this study proposed a novel self-supervised network, MOCO-MIM, for early ROP grading. The proposed classification network was evaluated on a total of 553 labeled fundus images from 89 preterm infants. The trained network achieved a test accuracy of 98.29% and an AUC score of 97.6% for three stages of grading. The adopted method is verified that the proposed method can be detected early stages of ROP more efficiently and grade the severity more accurately based on limited-labeled fundus images, which is superior to the existing state-of-the-art methods.

Feature extraction from the normalized transformation domain is a key technique in target detection. Traditional normalization approaches assume that matrix elements follow a normal distribution, but any deviations from this assumption can lead to significant systematic errors. This article presents a novel method that modifies the normalization process in the fractional Fourier transform (FRFT) domain by incorporating a threshold mechanism to counteract the effects of non-normal distributions. Three modified FRFT features are then extracted from this modified FRFT domain. Furthermore, we propose a target detection method that utilizes these three adjusted features. Experimental results based on measured data indicate that the modified FRFT features exhibit superior classification capabilities for sea clutter and targets compared to the original ones. Additionally, the experiments also demonstrate that under the same conditions, the proposed detection method outperforms traditional FRFT feature detector and the tri-feature based detector.

In a drone system with dual data links, this paper presents a redundant data processing algorithm that can minimize flight control instability without increasing the weight and price of the aircraft by software processing of duplicate received messages.

In this letter, we proposed and experimentally demonstrated a simple and efficient design method for high-voltage Floating Field Ring (FFR). Three floating FFRs is firstly used to determine the influence of rings’ distance (d) on the voltage drops of the highest voltage to second ring (Vn+1) and the second to the first ring (Vn). Then, two voltage curves (curve for Vn+1 and Vn) based on d are drawn. Then, for a target breakdown voltage, by iteration between the curves for Vn+1 and Vn, number of FFRs and corresponding distances can be found. Based on this method, an 8.0 kV PN diode with 110 FFRs is designed and fabricated. The measured breakdown voltage is 7840 V, which reaches 93% of the designed value. This method provides a simple and efficient design for high-voltage SiC devices, especially for ultra-high voltage applications where the number of rings exceeds dozens.

The introduction of YOLOv9, the latest version of the You Only Look Once (YOLO) series, has led to its widespread adoption across various scenarios. This paper is the first to apply the YOLOv9 algorithm model to the fracture detection task as computer-assisted diagnosis (CAD) to help radiologists and surgeons to interpret X-ray images. Specifically, this paper trained the model on the GRAZPEDWRI-DX dataset and extended the training set using data augmentation techniques to improve the model performance. Experimental results demonstrate that compared to the mAP 50-95 of the current state-of-the-art (SOTA) model, the YOLOv9 model increased the value from 42.16% to 43.73%, with an improvement of 3.7%. The implementation code is publicly available at https://github.com/RuiyangJu/YOLOv9-Fracture-Detection.

Introductions of quantum communications, enabled by advancements in quantum computing, is expected to play a significant role in the field of communications. Inherent properties of quantum objects, such as superposition and entanglement have the potential to provide novel solutions to overcome the challenges encountered by classical communication systems in bandwidth-intensive applications such as media transmission. This research explores the performance of a quantum communication system in image transmission using quantum superposition and investigates its performance using a simple quantum channel model. With increase of channel noise, there are significant gains in the rate distortion performance of images transmitted over the quantum channel, compared to an ideal classical channel. This novel attempt in constructing a quantum communication-based image transmission system indicates the potential of the approach to be applied to satisfy the ever-increasing demands of high-quality media transmission applications.

Occluded person Re-identification (Re-ID) is to identify a particular person when the person’s body parts are occluded. However, challenges remain in enhancing effective information representation and suppressing background clutter when considering occlusion scenes. In this paper, we propose a novel Attention Map-Driven Network (AMD-Net) for occluded person Re-ID. In AMD-Net, human parsing labels are introduced to supervise the generation of partial attention maps, while we suggest a Spatial-frequency Interaction Module (SIM) to complement the higher-order semantic information from the frequency domain. Furthermore, we propose a Taylor-inspired Feature Filter (TFF) for mitigating background disturbance and extracting fine-grained features. Moreover, we also design a part-soft triplet loss, which is robust to non-discriminative body partial features. Experimental results on Occluded-Duke, Occluded-Reid, Market-1501, and Duke-MTMC datasets show that our method outperforms existing state-of-the-art methods. The code is available at: https://github.com/ISCLab-Bistu/SA-ReID.

In this paper, based on thin film integrated passive device (TF-IPD) technology, compact ultra-wideband (UWB) bandpass filter (BPF) and dual-band bandpass filter (DBBPF) with equal inductance configuration are proposed. The configuration of equal inductance in a circuit contributes to reducing variables, simplifying the physical structure, and shortening the debugging time in the TF-IPD design process. Despite all the inductors being equal in value, wide passband can still be achieved. The measured results show that the center frequency of UBW BPF is 11.43 GHz, the 3-dB fractional bandwidth (FBW) is 133% (3.87 GHz-18.99 GHz). And the 3-dB FBWs of DBBPF centered at 6.5 GHz and 16.2 GHz are 92% (3.51-9.49 GHz) and 35% (13.33-18.98 GHz), respectively. The stopband of UBW BPF and DBBPF can be extended to 47.7 GHz and 44.8 GHz in sizes of 321 um × 856 um and 235 um × 846 um, respectively (excluding measuring pads).

The type parameter helps in the scattering mechanism analysis and scattering center identification. However, current approximate solution methods based on spectral estimation are noise-sensitive and have poor accuracy. In this letter, a high-precision approach for the type parameter based on the alternating direction method of multipliers (ADMM) is proposed. We use logarithmic transformation to separate the type parameter and the amplitude from the coupling term, and initially obtain the closed-form solution of the type parameter. Then, the regularization term of l2-norm is used for denoising. Finally, a joint optimization model based on ADMM is constructed to effectively estimate the type parameter. Simulation results confirm the high accuracy of the proposed approach.

For telecom operators, customer service is integral to their business. Traditional service systems, responsible for managing large amounts of data and complex knowledge bases, need more time retrieval processes and a lack of precision, hindering their ability to respond quickly to customer requests. To address these issues, this paper uses the LangChain programming framework to create a customized Large Language Model (LLM) specifically for the customer service context of telecom operators. It also uses reinforcement learning to improve the performance of the models and reduce the production of incorrect information. Experimental results show that the acceptance of our model’s recommended knowledge has increased from 15% to 70%, confirming its reliable operation in resource-constrained environments.

Edge computing has been proved an efficient approach to provisioning computation offloading service to vehicles on road through Road-Side Units (RSUs). However, the traffic volume on road is highly dynamic, while RSU-based edge servers are static in terms of geographical location and computation capacity. To address this problem, this paper proposes a mobile edge server placement strategy using cruising UAVs along the roads based on the genetic algorithm. We first build a mathematical model to characterize the deployment cost of these UAV-mounted servers and their routes. Next, we design a heuristic UAV-mounted edge server deployment scheme based on K-medoid clustering and genetic algorithms. Experimental results verify that our proposed UAV deployment scheme satisfies the offloading demand of IoV nodes while reducing the total deployment cost by 17.05% to 48.94% compared with existing popular approaches.

Targets with rotating components generate micro-motion (MM) modulation effect in addition to the main body. Extracting MM parameters is challenging due to interference from the target’s main body, necessitating the separation of modulation signals. This letter proposes a robust complex local mean decomposition (RCLMD) method with self-adaptive sifting stopping, aiming at the problem of component redundancy due to multiple iterations during break and the loss of modulation components during the separation process. The proposed method sets the objective function and self-adaptive stopping criterion, combined with the modulation signal characteristics, enhancing the accuracy and efficiency of MM component extraction. Simulation experiments indicate that at a low signal-to-noise ratio (SNR) of 3 dB, the separation effect of RCLMD is still 14.72\% higher than that of the conventional complex local mean decomposition (CLMD) method, and the separation efficiency is improved by 54.92\%. Furthermore, the measured radar signals verify the effectiveness of the proposed method in real scenarios.

In this letter, we focus on the problem of direct position determination (DPD) for moving targets. Compared with the traditional two-step localization methods, the DPD methods is more robust at low signal-to-noise ratio (SNR). However, to guarantee the optimal location results the computational complexity of DPD with grid search is too high, especially for moving targets. Therefore, we propose a new DPD method for moving targets with low computational complexity. First, using a proposed cost function, we obtain the position information from the received array signals directly. Second, we use the method of successional difference which averages the difference result of position over time to extract the velocity. It avoids the multi-dimensional parameters grid search and reduces the the computational complexity greatly. Simulation results demonstrate that the proposed algorithm outperforms other methods in a wide range of scenarios, especially at low SNR.

This paper presents a prediction-based dynamic range (DR) compensation technique in continuous-time (CT) delta -sigma modulator (DSM) with digital-domain noise coupling (DNC) structure. A digital prediction technique is proposed to address the issue of maximum stable amplitude (MSA) loss caused by DNC. The proposed prediction method has been shown to significantly extend the MSA of CT DSM with DNC. Moreover, through the extension of MSA, the DR and SNDR is significantly improved. A CTDSM with proposed compensation achieves a SNDR of 85.1dB at -1dBFS and a DR of 89dB.

Future communication networks are widely considered to be able to provide both high-speed communication service and reliable sensing service. Integrated sensing and communication (ISAC) has been a novel trend due to its possible hardware gain and spectrum gain by the integration of signal communication and sensing capabilities. Appropriate resource allocation strategy is essential for ensuring the quality of both communication and sensing services in terms of ISAC. Existing ISAC-based resource allocation schemes mainly focus on the coexistence of sensing and communication. However, in multi-user ISAC networks, the services required by users are diverse, including sensing-only, communication-only and dual-function cases. To this end, we proposes a resource allocation strategy based on user quality of service(QoS). Specifically, the sum rate of cellular network is maximized via optimizing spectrum resource selection and transmit power, while meeting the sensing QoS. We achieve the optimal solution to the entire problem by first demonstrating that the communication rate in this paper is a monotonically increasing function of sensing power, allowing us to obtain the optimal power allocation scheme. Subsequently, we employ a matching method to obtain the optimal spectrum sharing scheme.