Improving the downlink throughput of multi-UAV air-to-ground mmWave systems under rapid motion and channel uncertainty is a critical challenge. In this letter, UAV states are estimated and the resulting position uncertainty is propagated to the channel statistics through closed-form Jacobians. We propose an alternating optimization mechanism, where a surrogate sum-rate metric is used for inner-layer precoder design and a Riemannian trust-region solver is employed for outer-layer BS reconfiguration. Simulation results demonstrate downlink sum-rate improvements over non-optimized and Euclidean optimization baselines in dynamic multi-UAV mmWave deployments.

Infrared and visible image fusion aims to integrate complementary information from source images to generate high-quality fusion images that serve downstream tasks. However, the differentiated representation of image scene content, the unpredictability of degradation modes in source images, and the complexity of composite degradations pose significant challenges to constructing degradation-robust fusion models. To overcome these challenges, this study proposes STAFuse, a framework designed for composite degradation-robust image fusion that utilizes adaptive degradation-mode identification and aggregated textual prior guidance. We first introduce a scene- and degradation-aware mechanism that extracts crucial context and degradation data, converting it into a textual format to generate dynamic convolution kernels. This allows for the adaptive identification and elimination of varying degradation effects and scene disparities. Additionally, we implement an aggregated scene prior that condenses multi-source information into a fusion text, simulating an ideal scene to effectively guide the retrieval and fusion of multimodal information. Finally, we design a textual-domain supervision loss to perform auxiliary semantic supervision on the fusion model, thereby suppressing degradation effects and improving the visual fidelity of the fusion results. Experimental results demonstrate that the proposed method significantly outperforms existing state-of-the-art methods in terms of robustness, flexibility, and the aggregation of complementary information.

The energy-efficient transmission schemes are crucial to realize the Energy Harvesting (EH)-based Device-to-Device (D2D) communications. Multicast, one of the D2D modes, can serve as an effective approach to address the unreliable energy supply of EH-D2D communications and can further improve energy efficiency through cooperation among multiple users, but it has been rarely explored. To achieve the robust and energy-efficient performance for EH-D2D Multicast communications (EH-D2MD), we first design two cooperative transmission schemes: multi-cluster head content cooperation and single-cluster head energy cooperation by integrating the features of D2MD mode, efficient energy management method and wireless power transfer technology. To investigate the effectiveness and adaptability of the two cooperative schemes, we formulate a long-term average energy-efficient utility problem, which allocate the cluster heads, cooperative time and transmission power simultaneously and adaptively. We then propose an Online Convex Approximation (OCA) algorithm that combines the Lyapunov and convex approximation methods to address the non-convex Mixed Integer NonLinear Programming (MINLP) property of the modeled problem. With OCA, we can convert the long-term non-convex MINLP problem into a real-time convex MINLP problem, and obtain an optimal solution for this problem. Results reveal that the achieved energy efficiency of two proposed schemes is at least 10 times higher than that of no cooperation method, and improves at least 50% and up to 4 times compared to the single-slot cooperative algorithms.

Unmanned Aerial Vehicle (UAV) swarm applications, which are highly dependent on location information for task execution, encounter significant limitations when GNSS signal is denied under some harsh scenarios, such as military conflict environment. In such scenarios, unknown location information will lead to the failure of path planning, significantly impact task efficiency, and pose substantial challenges to location-based communication technologies. To address the above limitations, this paper proposes a Hierarchical Clustered Relative Localization Algorithm (DH-RLA). Based on the topological structure of nodes, the algorithm hierarchically partitions UAV nodes operating in GNSS-denied environments and adopts Multilateration Localization Algorithm (MLA) and Multi-Dimensional Scaling (MDS)-MAP algorithm for outer and inner layer localization, respectively. To ensure the accuracy and increase the speed of localization, a connectivity-enhanced anchor-assisted clustering algorithm is introduced to cluster the inner-layer nodes, enabling parallel execution of the corresponding localization algorithm within each cluster. Simulation results demonstrate that the proposed DH-RLA improves localization accuracy by 44% and 22.22% compared to MLA and MDS-MAP algorithm.

This paper introduces ns3-uavlink, a custom module designed to enable AI-driven dynamic Modulation and Coding Scheme (MCS) scheduling for Unmanned Aerial Vehicle (UAV) sidelink communications. Leveraging principles from Vehicle-to-Everything (V2X) standards, ns3-uavlink specifically targets UAV-to-UAV (U2U) scenarios. It enhances communication efficiency and reliability by adaptively adjusting MCS parameters under a fixed Block Error Rate (BLER) constraint. Implemented within the Network Simulator 3 (ns-3), ns3-uavlink offers a standardized environment for developing and evaluating Artificial Intelligence (AI) algorithms through a Python interface (pyuavlink), which uses efficient shared-memory data exchange mechanisms and supports the OpenAI Gym framework. Experimental evaluations demonstrate that the proposed adaptive scheme significantly improves system throughput compared to conventional fixed-MCS methods while maintaining a stable BLER. Overall, ns3-uavlink helps reduce the development cycle for AI-based UAV scheduling by streamlining development and accelerating prototyping within a unified UAV communication simulation framework.

In the coming 6G area, the throughput burden of cellular networks is generated by the the explosive growth in the number of smart devices and its requirements. The throughput burden can be mitigated effectively by the Device-to-Device Mul-ticast (D2MD) communication technology-based content sharing. With the aim of gaining the benefit of content sharing of D2MD, it is necessary to adopt an appropriate method for cluster formation that considers adaptivity and comprehensive influence. Thus, to address this method, this paper designs a global rate maximization problem by comprehensively taking into account the impact factors of the physical and social layers. Due to the high complexity of the designed problem, an Evolution-Affinity Propagation (EAP) algorithm is proposed to solve it. Finally, the performance of EAP is verified in simulation results and shows that the performance of EAP is improved by 2.5 times compared with spectral clustering.

The fluctuant and unreliable available energy has been the main reason for limiting the development of Energy Harvesting (EH)-based Device-to-Device (D2D) Communication. Multicasting can further improve energy efficiency by content sharing for EH-D2D communication. However, to satisfy the transmission Quality of Service (QoS) of cluster member with the worst channel state quality, the single cluster head in an EH-based D2D Multicast (EH-D2MD) Communication will face the conflict between energy shortage and consumption due to the requirement of energy consumption of multicasting and the fluctuation of available energy of EH. As a result, the QoS of multicasting transmission will deteriorate rapidly. To address the conflict, this paper designs an Adaptive Content Sharing scheme of Multi-cluster Head (ACS-MH) with the aim of improving the throughput of EH-D2MD. To solve the non-convex Mixed Integer Non-Linear Programming (MINLP) property of the throughput maximization problem and give it a union solution, we propose a three-step Convex Approximate lower-bound Algorithm (3-CDA). Simulations show that ACS-MH can effectively improve the throughput and spectrum efficiency of EH-D2MD.

Abstract Intelligent reflecting surface (IRS) has recently been considered as a potential technology for realizing ultra‐reliable and low‐latency (URLLC) in wireless networks. This paper proposes a resource optimization scheme to maximize the sum‐rate for an IRS‐assisted downlink multiuser multi‐input single‐output (MISO) URLLC system. For the perfect CSI scenario, we jointly optimize each user's block‐length and packet‐error probability, the precoding vectors at the base station (BS), and the passive beamforming with discrete phase shifts at the IRS. Given the problem's complexity, we design a computationally efficient iterative algorithm using successive convex approximation (SCA) and semidefinite relaxation (SDR) techniques to obtain a locally optimal solution. Specifically, for the imperfect CSI scenario, we construct a robust resource optimization problem model and incorporate the S‐procedure to address the impact of channel uncertainty, proposing an iterative algorithm based on the alternating optimization (AO) method to achieve a locally optimal solution. Simulation results demonstrate that: 1) An IRS equipped with a 2‐bit quantized resolution phase shifter is sufficient to achieve a system sum‐rate comparable to that of an ideal phase shifter; 2) Compared to other Baseline schemes, Algorithm 2 exhibits better robustness and superior performance gains under imperfect CSI.

UAV information dissemination or multicast is a critical use case for UAV communication systems. However, UAVs’ size and load limitations limit their onboard energy and endurance time. In this paper, we consider a multi-antenna unmanned aerial vehicle (UAV) enabled multicasting system, where a multi-antenna UAV is dispatched to transmit a shared file to a set of ground terminals (GTs). Our objective is to minimize mission completion time while satisfying the UAV speed and transmission power constraints, as well as guaranteeing the successful file recovery probability at the GTs. To address this issue, we introduce rank-two multicast beamforming and propose an online optimization scheme. The proposed scheme consists of two steps: 1) Before the UAV takes off, the open-loop rank-two beamforming without the channel state information (CSI) is used to obtain the initial planning, including the horizontal auxiliary distance parameter, UAV trajectory, UAV multicast trajectory segments (MTSs) on which the UAV will transmit multicast information, the number of multicasting packets to be transmitted at each MTS and mission completion time which is the upper bound. 2) Based on step 1, the UAV optimizes the rank-two multicast beamforming vectors using real-time CSI during flight to reduce the number of multicasting packets that need to be transmitted at each MTS, further decreasing mission completion time. Finally, numerical results indicate that the proposed scheme can significantly reduce mission completion time compared to the single antenna scheme.

In the field of infrared and visible image fusion, current algorithms often focus on complex feature extraction and sophisticated fusion mechanisms, ignoring the issues of information redundancy and feature imbalance. These limit effective information aggregation. To address these issues, this paper proposes a hierarchical fusion strategy with a two-stage training network, abbreviated as HF2TNet, which achieves effective information aggregation in a staged manner. In the initial training stage, a three-stream encoder-decoder architecture is proposed, seamlessly integrating CNN and transformer modules. This architecture extracts both global and local features from visible and infrared images, capturing their shared attributes before the fusion process. Moreover, a multi-shared attention module (MSAM) is proposed to profoundly reconstruct and augment the visible and infrared features, ensuring the preservation and enhancement of details across modalities. In the subsequent stage, HF2TNet utilizes the pre-integrated features as query inputs for the dual MSAMs. These modules interact with the previously reconstructed infrared and visible features to enhance complementary information and ensure a balanced feature fusion. Experimental results indicate HF2TNet's superior performance on standard datasets like MSRS and TNO, especially in complex scenes, demonstrating its potential in multimodal image fusion.

This paper investigates the block error rate (BLER) performance and proposes an optimal power allocation scheme of a two-hop decode-and-forward full-duplex relaying system with imperfect channel estimation information in a factory automation scenario that requires ultra-reliable low-latency communication (URLLC). In the relay-assisted system, a central controller S sends signals to a remote device D with the aid of a robot R with channel estimation error (CEE). First, considering CEE at each receiver and the residual loop interference of the full-duplex relay, we derive an approximate closed-form expression of BLER under the Nakagami-m fading channel. Next, to minimize the overall BLER, we propose an optimal power allocation scheme between the controller and the robot with a total power constraint. Furthermore, we prove that BLERs of S-R and R-D links are strictly convex in terms of the transmit power of the controller. Simulation results illustrate the performance of the considered relay-assisted system and validate our theoretical analysis.

Average-consensus protocol is one of the ways to develop distributed time-synchronization algorithms in Internet of Things (IoT) networks. However, the large number of iteration leads to a common time notion issue in nodes. This poses a critical challenge in the convergence of the time-synchronization algorithm and resulting asymptotic convergence in the average consensus protocol. In this article, a parameter-sharing-based average-consensus time-synchronization (PACTS) algorithm is proposed. For fast convergence, the proposed PACTS quickly forwards the time information to multihop nodes and employs multihop average-consensus instead of single-hop average consensus. Specifically, a node asynchronously and periodically broadcasts the relative clock offset estimation of neighbors with its local time information. Meanwhile, the relative clock offset estimation of the multihop node is calculated and used to estimate the average value. Consequently, an average consensus among local multihop nodes is obtained. As a result, the iteration number and convergence time are significantly reduced over the network. Finally, the experimental results indicate that the proposed PACTS algorithm has low complexity, high accuracy, and quick convergence.

In the beyond 5G/6G era, aerial edge computing (AEC) is expected to be used as significant components in Internet of Things. AEC brings flexible deployment with Line-of-Sight communication links for task offloading and computing services. In this article, we present a comprehensive survey of the AEC technology. First, we introduce a three-layer architecture of AEC, which includes the satellite, unmanned aircraft vehicles, and ground terminals. Second, we illustrate challenges in AEC, and summarize recent studies in terms of AEC performance metrics that include energy efficiency, latency, and operation cost to address challenges in AEC. Further, we introduce the advanced technologies applied in AEC management, e.g., artificial intelligence (AI) and the distributed optimization. Finally, the applications and the open issues with AEC are summarized and classified in the study.

Recently, Energy Harvesting-supported Device-to-Device (EH-D2D) communication receives extensive concerns due to its excellent properties in Energy Efficiency (EE), offloading capability, etc. Researches on EH-D2D mainly focus on resource allocation schemes of available energy. With the popularity of mobile devices and applications, content sharing becomes anytime, anywhere. Multicast is the foundation to realize the content sharing and can reduce transmission consumption because of multi-user service feature. Thus, Mutlicast can further improve EE of EH-D2D in communication mode rather than in resource allocation. However, the heterogeneity and imbalance of available energy, content requests and social ties on the user side will bring great challenges to the deployment of EH-D2D Multicast (EH-D2MD) communication. This paper aims at EE optimization and establishes a two-layer cluster formation model, which solves grouping, cluster head selection and power control jointly. When facing the non-convex of the sum of fractional functions and the product of binary variables, this paper specifically designs a two-step Convex Approximation Algorithm (CDA). CDA skillfully transforms the modeled non-convex Mixed Integer Non-Linear Programming (MINLP) problem into a convex MINLP one, which is easier to solve. Simulation results reveal that CDA can obtain the joint cluster formation results with approximate optimal EE and lower complexity.

Radio-frequency-energy-powered cognitive radio network (RF-CRN) is being taken seriously in Connected Vehicles, especially in 5G network, which can better address the challenges of energy limitation and spectrum scarcity. However, the energy efficiency (EE) of the RF-CRN wherein multiple secondary users (SUs) share the same channel is rarely presented. In this article, we consider a RF-CRN in which SUs first harvest energy from RF signals originating from a primary network (PN) and then utilize the available energy in the battery to transmit data. Since all SUs can access the authorized spectrum for transmission simultaneously, co-frequency interference (Co-FI) occurs among SUs. Given the quality of service (QoS) requirement, our goal is to achieve the maximum EE of the RF-CRN by jointly optimizing transmission time and power control. To this end, a resource allocation scheme referred to as approximate convex policy for co-frequency interference (CO-ACP) is proposed. Specifically, the EE problem is firstly converted into a convex one by CO-ACP. Then, we utilize Frank-Wolfe (FW) and one-dimensional linear programming to obtain the optimal solution. Simulation results demonstrate that a tight lower-bound optimum solution for the non-convex EE maximization can be achieved by CO-ACP. Moreover, the CO-ACP provides meaningful system features, such as the number of SUs, energy harvesting efficiency, and the battery energy state of the SUs under different RF-CRN scenarios, providing a clear reference for future deployment of RF-CRN.

Efficient energy management, especially in terms of energy efficiency (EE) optimization, is the cornerstone to realize the energy harvesting (EH)-powered device-to-device (D2D) communication underlaying cellular network. However, too much focus on energy management is likely to negatively impact other quality-of-service, such as transmission delay. Consequently, this paper aims to investigate an EE and delay tradeoff (EDT) optimization scheme in EH-powered D2D communication underlaying cellular network (EH-DCCN). A more practical scenario, one cellular user and multiple EH-powered D2D pairs matching model and limited energy and data storage space, is considered. In view of the characteristic of the available transmission energy, a weighted EE and delay utility problem is modeled to investigate the EDT optimization scheme. Due to the complexity of the modeled problem, this paper proposes a modified distributed Q-learning (QL) algorithm, namely experience-sharing distributed cooperation learning (EDCL) algorithm, to tackle the EDT optimization issue and enhance the convergence speed. Numerical results discuss convergence speed of EDCL performance, proper experience-sharing interval setting, as well as EDT performance. With the proper experience-sharing interval, EDCL can obtain the near-optimal performance as the classical centralized QL mechanism with more rapid convergence rate by sacrificing the appropriate tolerable additional signaling overhead.

Hard energy constraints impose huge challenges to wireless communication under the constraint of communication delay. In this paper, we consider the problem of energy-efficient point-to-point transmission scheduling of delay sensitive multimedia data over a fading channel. We first formulate the transmission scheduling problem as a Markov Decision Process (MDP) and systematically unravel it by the optimal solution. We then propose a Heuristic Evaluation Post-decision State (HE-PDS) learning algorithm for the problem of energy-efficient scheduling with respect to the communication delay. This algorithm is based on reformulating the value iteration equation by introducing a virtual state called Post-decision State (PDS). The advantages of the proposed algorithm are that: (i) it exploits only part of the information about the system so that less information needs to be learned than when using conventional reinforcement learning algorithms; (ii) it uses the heuristic function and evaluation function to reduce unnecessary action exploration in the whole search space, which severely limits the adaptation speed and runtime performance of conventional reinforcement learning algorithms; (iii) it chooses actions with favorable purpose and efficiency so as to optimize reward function and enhance convergence speed. We compare our algorithm with the traditional Q learning algorithm and PDS learning algorithm in terms of the energy consumption and transmission delay. The simulation results illustrate the performance of the proposed algorithm under various scenarios achieves a better trade-off between energy consumption and delay overhead. Moreover, it can converge in a reasonable number of slots for it to be practically useful with satisfaction of delay constraint.

The gateway location register (GLR) in the Universal Mobile Telecommunication System (UMTS) core network is a national node between the VLR and/or SGSN and the HLR. It handles location management created by roaming users in visited network, usually located far away from users' home network, without involving the HLR in every change of location areas (LAs). However, when the number of roaming subscribers increases in a visited network, the centralized GLR may become a bottleneck of the system, which results in degrading system performance and quality of service. In this paper, we present a novel distributed GLR scheme for 3G cellular networks where the location management signaling overhead is evenly distributed to each VLR in the visited network. Analytic results show that our distributed scheme outperforms the traditional centralized GLR scheme. In addition, the system's robustness is greatly enhanced in our proposed system