Recently, unmanned aerial vehicles (UAVs) are crucial in the low-altitude economy due to their high mobility and operational efficiency, which enable rapid and flexible operations in various applications. To support the flight control and management of low-altitude UAVs, ground base stations need to ensure communication quality that is high-capacity, ultra-low latency, and highly reliable. Extremely large-scale multiple-input multiple-output (XL-MIMO) systems, capable of forming high-gain directional beams, are highly suited for supporting the communication requirements of UAVs in the low-altitude economy. However, the communication signal may quickly deviate from the main lobe of the beam due to the mobility of UAVs, resulting in beam misalignment and significant degradation in communication quality. To address this problem, this paper first analyzes the near-field beam pattern of XL-MIMO system serving low-altitude UAVs and then decomposes phase of beam pattern into linear and nonlinear components to derive the angular and distance half-power beamwidth in the near field. Then, the prediction error is used to determine the number of activated antennas for dynamic beamwidth adjustment based on UAV position prediction. In addition, a near-field adaptive beamwidth control-aided tracking (NF-ABCT) algorithm is proposed to improve the effectiveness of beam tracking. Finally, simulation results demonstrate that the proposed NF-ABCT achieves a higher signal-to-noise ratio compared to the existing fixed beam scheme and channel estimation methods, while maintaining beam traking robustness and stability in both near-field and far-field.

The ocean contains vast resources such as biological, mineral, and strategic resources for national defense. Developing underwater wireless sensor networks (UWSNs) to achieve collaborative data collection and transmission is essential for marine exploration activities. The Internet of Things (IoT), enabling real-time data transmission, holds immense potential in its application to UWSNs, known as the underwater Internet of Things (UIoT). Underwater vehicles (UVs), as mobile nodes in UIoT, play an irreplaceable role in enabling dynamic network topology and overcoming static node coverage limitations. However, in the face of massive data generated by UIoT devices, seeking an efficient communication method has become increasingly important. Underwater wireless optical communication (UWOC) technology has attracted extensive attention in underwater communication due to its high-speed data transmission capability and superior transmission capacity. Nevertheless, existing UWOC systems ignore the semantic information of transmitted data, leading to unnecessary communication resource overhead. Furthermore, the spatiotemporally dynamic marine environment presents significant challenges to UWOC technology. This paper integrates semantic communication into the challenging scenario of underwater wireless optical image transmission to further optimize the bandwidth compression efficiency and anti-interference capability of underwater optical wireless communication. In the experimental verification section, the results demonstrate that the communication scheme proposed in this paper can achieve excellent image transmission performance in variable seawater channels, outperforming the traditional source-channel separation communication paradigm in terms of image transmission performance and exhibiting certain robustness.

To address multi-user interference and codebook design challenges in sparse code multiple access (SCMA) systems, we propose a Wasserstein generative adversarial network (WGAN) integrated convolutional residual neural network (CR-Net) for end-to-end joint optimization of hybrid autoencoder-channel components. The proposed architecture applies convolutional neural networks (CNNs) for multidimensional codebooks encoding and residual neural networks (ResNet) for interferenc-resistant decoding, forming a hybrid autoencoder structure. In practical wireless channels with variable timing, encoder-generated codebooks do not dynamically adapt to channel variations, resulting in a degraded bit error rate (BER) performance. We employ WGAN to model channel effects: 1) the generator models channel distortion to drive encoder-generated channel-adaptive codebooks; 2) the discriminator uses Wasserstein distance-based feature extraction to co-optimize decoder interference suppression via backpropagation. Simulation results demonstrate BER performance and computational complexity that surpasses deep learning (DL)-based benchmarks.

As an essential technology for spectrum sensing and dynamic spectrum access, automatic modulation classification (AMC) is a critical step in intelligent wireless communication systems, aiming at automatically recognizing the modulation schemes of received signals. In practice, AMC is challenging due to the influence of communication environment and signal parameters, such as unknown channels, noise, symbol rate, signal length, and sampling frequency. In this survey, we investigated a series of typical AMC methods, including key technology, performance comparisons, advantages, challenges, and future key development directions. According to the methodology and processing flow, AMC methods are divided into three categories: likelihood‐based (Lb) methods, feature‐based (Fb) methods, and deep learning methods. The technical details of various types of methods are introduced and discussed, such as likelihood distributions, artificial features, classifiers, and network structures. Then, extensive experimental results of state‐of‐the‐art AMC methods on public or simulated datasets are compared and analyzed. Despite the achievements that have been made, there are still limitations of the individual methods, including generalization capability, reasoning efficiency, model complexity, and robustness. In the end, we summarized the severe challenges faced by AMC and key future research directions.

With the wide application of deep learning (DL) across various fields, deep joint source–channel coding (DeepJSCC) schemes have emerged as a new coding approach for image transmission. Compared with traditional separated source and CC (SSCC) schemes, DeepJSCC is more robust to the channel environment. To address the limited sensing capability of individual devices, distributed cooperative transmission is implemented among edge devices. However, this approach significantly increases communication overhead. In addition, existing distributed DeepJSCC schemes primarily focus on specific tasks, such as classification or data recovery. In this paper, we explore the wireless semantic image collaborative nonorthogonal transmission for distributed edge networks, where edge devices distributed across the network extract features of the same target image from different viewpoints and transmit these features to an edge server. A two‐view distributed cooperative DeepJSCC (two‐view‐DC‐DeepJSCC) with or without information disentanglement scheme is proposed. In particular, the two‐view‐DC‐DeepJSCC with information disentanglement (two‐view‐DC‐DeepJSCC‐D) is proposed for achieving balancing performance between multitasking of image semantic communication; while the two‐view‐DC‐DeepJSCC without information disentanglement only pursues outstanding data recovery performance. Through curriculum learning (CL), the proposed two‐view‐DC‐DeepJSCC‐D effectively captures both common and private information from two‐view data. The edge server uses the received information to accomplish tasks such as image recovery, classification, and clustering. The experimental results demonstrate that our proposed two‐view‐DC‐DeepJSCC‐D scheme is capable of simultaneously performing image recovery, classification, and clustering tasks. In addition, the proposed two‐view‐DC‐DeepJSCC has better recovery performance compared to the existing schemes, while the proposed two‐view‐DC‐DeepJSCC‐D not only maintains a competitive advantage in image recovery but also has a significant improvement in classification and clustering accuracy. However, the proposed two‐view‐DC‐DeepJSCC‐D will sacrifice some image recovery performance to balance multiple tasks. Furthermore, two‐view‐DC‐DeepJSCC‐D exhibits stronger robustness across various signal‐to‐noise ratios.

In highly interconnected large-scale event and other Internet of Things (IoT) device-intensive scenarios, traditional terrestrial base stations have difficulty meeting the requirements of IoT devices for network speed and security, and have exacerbated carbon pollution. To this end, a blockchain-enabled unmanned aerial vehicles (UAVs)-assisted mobile edge computing (MEC) system is introduced to enhance communication efficiency and ensure the privacy of IoT devices. In this system, the Byzantine consensus algorithm is applied in the blockchain. Considering the pollution of reducing carbon dioxide emissions, a strategy for jointly optimizing the flight trajectories of UAVs, task offloading scheduling, and MEC computing resource allocation is formulated to minimize the system’s carbon emissions and time delay while meeting MEC and blockchain computing tasks. However, due to the coupling of variables, this problem is very complex. Therefore, the original problem is decoupled into multiple subproblems, and the block coordinate descent method (BCD) and successive convex approximation method (SCA) are used for solving. Specifically, the UAV flight trajectories, task offloading scheduling, and MEC computing resource allocation are alternately optimized until convergence. Simulation results verify the effectiveness and good performance of the proposed algorithm in this article.

This article analyses the performance of a power-domain nonorthogonal multiple access (PD-NOMA)-based visible light communication (VLC) system in the Internet of Things (IoT) network. It specifically investigates the various VLC system channel models in indoor and outdoor IoT network settings. To enhance the existing analysis of the PD-NOMA-based VLC system performance, this study first analyses the distribution function of the received signals and the system noise in real-world environments. Second, it explores the impact of environmental noise on the system’s bit error rate (BER) under indoor Gaussian channels and derives the theoretical BER formulas for both the indoor and outdoor settings. Finally, theoretical symbol error rate (SER) formulas are derived for the three-device PD-NOMA-VLC system in both the indoor and outdoor environments. Numerical simulations demonstrate a close match between the theoretical analysis and simulation results, particularly in high signal-to-noise ratio (SNR) conditions.

Cell instance segmentation (CIS) via light microscopy and artificial intelligence (AI) is essential to cell and gene therapy-based health care management, which offers the hope of revolutionary health care. An effective CIS method can help clinicians to diagnose neurological disorders and quantify how well these deadly disorders respond to treatment. To address the CIS task challenged by dataset characteristics such as irregular morphology, variation in sizes, cell adhesion, and obscure contours, we propose a novel deep learning model named CellT-Net to actualize effective cell instance segmentation. In particular, the Swin transformer (Swin-T) is used as the basic model to construct the CellT-Net backbone, as the self-attention mechanism can adaptively focus on useful image regions while suppressing irrelevant background information. Moreover, CellT-Net incorporating Swin-T constructs a hierarchical representation and generates multi-scale feature maps that are suitable for detecting and segmenting cells at different scales. A novel composite style named cross-level composition (CLC) is proposed to build composite connections between identical Swin-T models in the CellT-Net backbone and generate more representational features. The earth mover's distance (EMD) loss and binary cross entropy loss are used to train CellT-Net and actualize the precise segmentation of overlapped cells. The LiveCELL and Sartorius datasets are utilized to validate the model effectiveness, and the results demonstrate that CellT-Net can achieve better model performance for dealing with the challenges arising from the characteristics of cell datasets than state-of-the-art models.

Abstract Skin cancer, also known as melanoma, is a deadly form of skin cancer that can significantly improve survival rates when diagnosed at an early stage. It is usually diagnosed visually from dermoscopic images, and such visual assessment of skin cancer by the naked eye is a challenging and arduous task. Therefore, the detection of melanoma from dermoscopic images using trained artificial intelligence models is of great importance today. However, since melanoma is a rare disease, existing databases of skin lesions often contain highly unbalanced numbers of benign and malignant samples. In this paper, we propose a new one‐class classification‐based skin lesion classification strategy for small and unbalanced datasets. One‐class classification (OCC) is a special case of multi‐classification. OCC aims to learn a descriptive paradigm from positive class data (true data) during training and reject pseudo data (fake data) that do not conform to the paradigm during inference. OCC has great potential for application in anomaly detection problems. We have analyzed several approaches to the OCC task in recent years and propose a new design paradigm for the OCC problem, taking into account the unbalanced data set of the melanoma classification task. We have designed an improved OCC network based on this design paradigm, where the network is based on the architecture of a residual neural network, combining the coding and decoding idea of variational self‐encoder and the adversarial training idea of an adversarial neural network, using binary cross‐entropy as the loss function and introducing the channel attention mechanism. Tests on several publicly available dermatology datasets show that this improved OCC network addresses the unbalanced dataset situation in melanoma image classification to some extent while having relatively excellent performance. Compared with some traditional networks, it can obtain more stable training results and perform more consistently on complex datasets.

Abstract Unmanned aerial vehicles (UAVs) are widely used in wireless communication networks due to their rapid deployment and high mobility. However, in practical scenarios, the existence of obstacles and eavesdroppers will seriously interfere with the communication quality of the UAV network and produce a security risk. Thus, this paper combines reconfigurable intelligent surface (RIS) technology with UAVs to build a secure UAV communication network. Normally, a rotary‐wing UAV (labeled as UAV‐S) acting as a base station sends information signals to a legitimate user on the ground with RIS equipment. However, there is a passive eavesdropper on the ground who can steal the information. Therefore, a friendly UAV jammer (labeled as UAV‐J) with a fixed location is introduced to send jamming signals to confuse the eavesdropper. The goal of this paper is to maximize the average secrecy rate of the communication network by jointly optimizing the flight trajectory, transmit power of the UAV‐S and UAV‐J, and phase shifter of the RIS. Since the constructed problem is highly nonconvex, an alternating optimization algorithm based on successive convex approximation techniques is proposed to solve the problem. Simulation results show that the proposed algorithm can achieve a higher secrecy rate in comparison with other schemes.

Abstract This paper studies a multi‐unmanned aerial vehicle (UAV) assisted mobile edge computing (MEC) system, where multiple UAVs equipped with MEC server are deployed to provide computing offloading services for ground users with limited local computing resources. Specifically, each user divides its computing tasks into two parts: one part is offloaded to the associated UAV for calculation, and the remaining part is calculated locally. It is aimed to minimize the weighted energy consumption of all UAVs and all users by jointly optimizing the UAV trajectory, user scheduling, central processing unit (CPU) calculation frequency allocation, and offloading time allocation. However, the formulated problem is a mixed integer non‐convex problem, which is challenging to obtain an optimal solution. In order to effectively solve this, an efficient two‐stage iterative algorithm based on alternative optimization is proposed by decomposing the original problem into two sub‐problems to obtain a suboptimal solution. The simulation results show that the proposed scheme is superior to other benchmark schemes in reducing system energy consumption.

With the explosive growth of ubiquitous mobile services and the advent of the 5G era, ultra-dense wireless network (UDN) architectures have entered daily production and life. However, the massive access capacity provided by 5G networks and the dense deployment of micro base stations also bring challenges such as high energy consumption, high maintenance costs, and inflexibility. Fiber-based visible light communication (FVLC) has the advantages of large bandwidth and high speed, which provides an efficient connection option for UDN. Thus, in order to make up for the poor flexibility of UDN, we propose a new FVLC-UDN architecture based on software-defined networks (SDNs). Specifically, SDN decouples the data plane and the control plane of the device and centralizes the control of the LED in the cell through a unified control plane, which can not only improve the resource allocation ability of the network but also transmit the data only as the data plane, reducing the manufacturing and implementation costs of the LED. In order to get a better resource allocation scheme, this paper proposes a model for predicting cell traffic based on convolutional neural networks. By predicting the traffic of each cell in the control domain, the traffic trend and cells’ status in the future period of time in the control domain can be obtained, so that a much more efficient resource allocation scheme can be formulated proactively to reduce energy consumption and balance communication loads. The experimental results show that on the real cell traffic dataset, this method is better than the existing prediction methods when the size of training dataset is limited.

In this paper, the energy efficient virtual machine (VM) placement and virtual resource allocation problem for the mobile edge computing (MEC) system is explored. Particularly, we develop an optimization framework of energy consumption minimization for computing and offloading by jointly optimizing the VM placement matrix and the number of physical machines (PMs). To resolve this problem, we transform the optimization problem into a non-uniform weighted hypergraph model. In this model, the weight of hyperedge is defined as the negative of the accumulated energy consumption for computing at VM instances hosted by one PM. Based on the hypergraph model, a hypergraph matching algorithm by utilizing the local search policy is proposed for finding the maximum-weight subset of vertex-disjoint hyperedges, aiming to obtain an optimal VM placement, i.e., (M*)-perfect matching. Furthermore, the virtualized resources are further allocated to user equipments (UEs) in the form of multiple VM instances via the optimal VM placement to meet the requirement of workloads. Simulation results are presented to demonstrate the effectiveness of the proposed hypergraph matching algorithm over the alternative benchmark algorithm.

We study the optimum designs of the downlink of user-centric ultra-dense Internet of Things (IoT) networks with fiber-wireless communications (FWCs). A large number of low power radio access points (RAPs) are densely deployed in the network to provide service to spatially distributed IoT physical devices (PDs). The RAPs are connected to a central unit (CU) through optical fiber (OF) front-hauls. Radio-frequency-over-fiber (RFoF) is employed in the optical front-hauls to reduce RAP complexity, cost, and energy consumption. With RFoF front-hauls, wireless signals received by PDs are subject to distortions accumulated through the optical and wireless links, including optical loss, optical chromatic distortion, optical and thermal noises, wireless pathloss, and small scale fading. The optimum designs are performed across the optical and wireless domains with the help of a newly developed model that quantifies the combined effects of the optical and wireless links. One of the main challenges faced by the design of an ultra-dense IoT network is the high energy consumption due to dense RAP deployment. The objective of this paper is to minimize the total energy consumption of the entire IoT network, including both optical and wireless links, by jointly optimize RAP power allocation and RAP-PD association, subject to quality-of-service (QoS) constraints for each PD. We propose a low complexity suboptimum binary forcing gradient search (BFGS) algorithm, which performs a gradient-based search based on the unique structure of the problem. Simulation results show that the optical front-hauls have significant impacts on the performance and design of ultra-dense IoT networks.

In this paper, an optimized codebook design based on rotated angles and extrinsic information transfer (EXIT) chart for a non-orthogonal multiple access scheme, called irregular sparse code multiple access (IrSCMA), is presented. Unlike regular SCMA, in IrSCMA, the defined degree of layer's resource nodes is different, which is beneficial to system performance.It is demonstrated that the new codebook can greatly improve the BER performance especially when the signal-to-noise ratio(SNR) is high, without sacrificing the low detection complexity, compared with the existing codebooks and LDS.