In this paper, we study a cellular-connected unmanned aerial vehicle (UAV) which aims to fly between two predetermined locations while maintaining ultra-reliable low-latency communications (URLLC) for command-and-control (C2) links with terrestrial base stations (BSs). Long-range flights often trigger frequent inter-cell handovers, which may introduce delays and synchronization overhead. We jointly optimize the continuous trajectory and BS association to minimize handovers, path length, and flying time, subject to communication reliability and kinematic constraints. To address this problem, we reformulate it as an optimization based on the graph of convex sets (GCS). First, the URLLC requirement is translated into spatially feasible regions in the flight plane for each BS. And an intersection graph is constructed including the start and goal points. Each graph node is associated with a smooth and dynamically feasible trajectory segment. The trajectory is parameterized in space by Bézier curves and in time by a monotonic Bézier scaling, together with convex constraints that ensure continuity and enforce speed bounds. Next, we impose unit-flow constraints to enforce a single path, and by coupling the resulting binary edge-selection variables with the convex constraints, we obtain a mixed-integer convex program (MICP). Applying a convex relaxation and rounding to the mixed-integer convex program produces nearly globally optimal routes, and a final refinement yields smooth, dynamically feasible trajectories. Simulations verify that the method preserves URLLC connectivity while achieving a clear trade-off between fewer handovers and flight efficiency.

The Internet of Thing (IoT) devices play a vital role in collecting mission-critical and time-sensitive sensing data from remote areas, where traditional terrestrial networks are constrained by sparse infrastructures. However, resource-limited ground devices (GDs) in such scenarios often lack the ability to directly transmit essential information to distant data centers. To overcome this challenge, this paper proposes a Satellite-unmanned aerial vehicle (UAV)-assisted data collection framework, where the UAV is controlled by a remote control center via satellite relays. Aiming to maximize the energy efficiency (EE) of the UAV, we first design a reference trajectory to the UAV with given hovering positions. Subsequently, we optimize the power allocation for communication and state sensing strategies for trajectory tracking control, while guaranteeing control stability and communication reliability. These challenging problems are addressed using sequently an efficient algorithm, incorporating Deep Q-Network (DQN), closed-form derivations, and one-dimensional search method. Extensive numerical simulations and experimental validations are conducted to demonstrate the effectiveness of the proposed approach. Key findings point that the data size of collection has greater impacts than transmission power. Moreover, the results reveal the relationships among the communication, control and state sensing in terms of the EE.

Relative localization of unmanned aerial vehicle (UAV) swarms in GNSS-denied environments is critical for emergency rescue and battlefield reconnaissance. Existing methods face large errors from packet loss and high complexity in large swarms. This paper proposes a clustering-based framework where UAVs leverage communication signals for channel estimation and ranging. Spectral clustering first divides the swarm into sub-clusters, where matrix completion and multidimensional scaling recover relative coordinates. A global map is then formed through inter-cluster anchor fusion. A case study of UAV integrated communication and sensing (ISAC) system is presented, where the Orthogonal Time Frequency Space (OTFS) is adopted for ranging and communication. Experimental results show that the proposed method reduces localization errors in large swarms and loss of range information. It also explores the impact of signal parameters on communication and localization, highlighting the interplay between communication and localization performance.

We present MEGA-Bench, an evaluation suite that scales multimodal evaluation to over 500 real-world tasks, to address the highly heterogeneous daily use cases of end users. Our objective is to optimize for a set of high-quality data samples that cover a highly diverse and rich set of multimodal tasks, while enabling cost-effective and accurate model evaluation. In particular, we collected 505 realistic tasks encompassing over 8,000 samples from 16 expert annotators to extensively cover the multimodal task space. Instead of unifying these problems into standard multi-choice questions (like MMMU, MM-Bench, and MMT-Bench), we embrace a wide range of output formats like numbers, phrases, code, \LaTeX, coordinates, JSON, free-form, etc. To accommodate these formats, we developed over 40 metrics to evaluate these tasks. Unlike existing benchmarks, MEGA-Bench offers a fine-grained capability report across multiple dimensions (e.g., application, input type, output format, skill), allowing users to interact with and visualize model capabilities in depth. We evaluate a wide variety of frontier vision-language models on MEGA-Bench to understand their capabilities across these dimensions.

In this paper, we present a remote UAV control framework to achieve the tasks of trajectory tracking while avoiding collision. We utilize short-packet communication to transmit the control commands. Following this framework, we develop an improved self-triggered stochastic model predictive control (ST-SMPC) to adaptively adjust the control length, which greatly affects the control and communication performance. Moreover, chance constraints are considered to realize collision avoidance. Consequently, we decouple this co-design problem into a SMPC problem and a communication power optimization problem to jointly determine the control length. Simulation results demonstrate that the proposed control method can adaptively adjust the control length under various environments, achieving better communication efficiency compared to traditional MPC approaches.

In this paper, a method combining event triggered control (ETC) and packetized predictive control (PPC) are proposed for remote UAV trajectory tracking. In particular, we utilize PPC to reduce the communication consumption in wireless networks while ensuring robustness. The ETC is adopted to reduce the number of commands required for control while ensuring the control stability. A new index, reflecting the communication energy per sampled time, is given and minimized by optimizing communication power and control command packet length jointly. Simulations are carried out to showcase the effectiveness of our method in tracking performance and communication cost. Compared with periodic PPC methods, the proposed method can reduce communication power consumption while achieving better control performance.

Recent advancements in end-to-end autonomous driving have primarily relied on data-driven approaches. However, these methods face challenges regarding interpretability and safety guarantees in motion prediction results. We draw inspiration from the knowledge-driven paradigms used in human driving to address these issues. Our exploration focuses on how to incorporate similar capabilities into autonomous driving (AD) systems. We propose LLMDriver, a Large Language Model (LLM)-based agent that integrates an interactive driving environment, multiple driving agents, and a memory module. This framework leverages the potential of large language models (LLMs) to develop a system that drives like humans. By reasoning through the driving actions taken and accumulating experiences from continuous driving, our approach aims to enhance motion planning in autonomous vehicles. LLMDriver achieves an L2 error of 0.82 and a 0.31% collision rate in UniAD metrics, performing comparably to UniAD and slightly below RDA Driver. Under ST-P3, it attains a$\text{0. 4 1}$L2 error and$\text{0. 1 1 \%}$collision rate. In CARLA closed-loop tests, it scores 65% in driving and achieves the highest route completion, surpassing prior methods by 1.1%.

This article presents a distributed optimal control method to achieve formation control with obstacle avoidance and connectivity maintenance of unmanned aerial vehicles (UAVs). Specifically, a control Lyapunov function (CLF) consisting of information from neighbors is constructed to guarantee the asymptotic stability of the UAV formation. Then, obstacle avoidance and connectivity maintenance are achieved using the control barrier function (CBF), where the UAV control has been considered under limited communication power constraints. Simulation experiments demonstrate that the proposed method can realize effective formation control and communication power allocation.

In next-generation wireless networks, the integration of localization and communications techniques are regarded as a paradigmatic shift for enhancing spectrum and hardware utilizations. The channel sensing, encompassing localization and channel estimation, plays a pivotal role in various aspects such as beamforming, precoding and high-quality data transmission. In this paper, we present a method for optimizing the frame structure in the localization and communication integration system, to reveal the intricate relationship among channel estimation, user's localization and communication throughput in terms of the spectral efficiency (SE). Specifically, we initially derive the error bounds for channel estimation and location prediction in dynamic point-to-point communication scenario. Leveraging these bounds, we optimize the sensing and communication duration together with beamwidth design, to maximize the SE while ensuring communication requirements. An efficient iterative algorithm is employed to tackle this non-convex problem. Numerical results demonstrate that our proposed method can achieve a nearoptimal SE performance with significantly lower complexity compared to exhaustive search method. Furthermore, our results underscore the critical role of localization in optimizing sensing and communication durations for SE, particularly in high dynamic scenarios.

Since the traffic administration at road intersections determines the capacity bottleneck of modern transportation systems, intelligent cooperative coordination for connected autonomous vehicles (CAVs) has shown to be an effective solution. In this paper, we try to formulate a Bi-Level CAVs intersection coordination framework, where coordinators from High and Low levels are tightly coupled. In the High-Level coordinator where vehicles from multiple roads are involved, we take various metrics including throughput, safety, fairness and comfort into consideration. Motivated by the time consuming space-time resource allocation framework, we try to give a low complexity solution by transforming the complicated original problem into a sequential linear programming one. Based on the “feasible tunnels” (FT) generated from the high-Level coordinator, we then propose a rapid gradient-based trajectory optimization strategy in the low-level planner, to effectively avoid collisions beyond high-level considerations, such as the unexpected pedestrian or bicycles. Simulation results and laboratory experiments show that our proposed method outperforms existing strategies. Moreover, the most impressive advantage is that the proposed strategy can plan vehicle trajectory in milliseconds, which is promising in real-world deployments. A detailed description include the coordination framework and experiment demo could be found at the supplement materials, or online at https://youtu.be/MuhjhKfNIOg.

In this paper, we propose a novel UAV aided ground nodes (GNs) localization and communication integrated framework, where the age of information (AoI) is introduced to evaluate the system timeliness. We aim to jointly optimize the UAV trajectory, localization accuracy, bandwidth and beamwidth, to guarantee the information freshness. Specifically, we give a two-stage method, where a low complexity initial UAV trajectory searching algorithm is firstly proposed, based on theroughposition information of GNs. Afterwards, we formulate a joint UAV location and resource optimization problem. This essential mixed integer problem can be solved by efficient successive convex approximation based iterative algorithm. Simulations show that the localization and communication integrated framework can obtain about 50% performance gain compared with the UAV communication aid only solution, and over 37% performance gain via proper resource allocation. Moreover, the analysis reveals that our proposed scheme strikes a balance among the time durations of localization, data transmission and UAV movement. Additionally, we conduct practical experiments to draw valuable insights into the system design and implementations (An experimental can be found on the supplementary materials, or online at https://youtu.be/OX6Bgz6naUA).

A crucial aspect of human-robot integration is the implementation of person-following robots. However, autonomous robots continue to face challenges in tracking individuals within complex environments with dynamic obstacles. Traditional approaches to combining global and local planning are inadequate for this task due to the high level of uncertainty in the environment and the flexible behavior of the following target. To address these challenges, this paper proposes a framework for motion planning for wheeled robots. This method integrates dynamic obstacle avoidance and dynamic target following into an optimization problem based on model predictive control (MPC) with terminal constraint set to guarantee the safety of the following task in dynamic obstacle environments. Therefore, in this paper, a high-frequency state estimator is used to predict human behavior, and a spring model is used to model dynamic obstacles in the environment to keep the robot away from the obstacles. The effect of the person following is thoroughly tested in simulation with multiple scenarios and the comparison experiment, verifying the real-time effectiveness of the method.

Location-based beamforming has emerged as a promising approach for both complexity reduction and beam alignment. However, most existing solutions only perform well when the position information is accurate, which is challenging in highly dynamic scenarios with frequent location changes. This paper introduces a novel three-phase location-based beamforming framework that is designed to exhibit robustness against potential position uncertainty. The framework employs Bayesian filtering to tackle the non-Gaussian process in the location prediction phase. A robust optimization problem is then formulated to generate a cubic-like beam, ensuring the quality of service even at the beam edge. Simulation results validate the robustness of our beamforming scheme, demonstrating its ability to maintain stable links and deliver satisfactory channel rate.

The traditional Global Navigation Satellite System (GNSS)-based localization and tracking method will lead to a notable performance reduction when the navigation signals are obstructed. In this paper, by utilizing the inherent benefits of UAVs, such as reliable line-of-sight (LoS) propagation and exceptional mobility, we employ the UAVs to achieve the seamless localization and tracking. Specifically, we first analytically derive distribution of noise for the vehicle-to-UAV (V2U) measurements. Subsequently, we provide a general distributed simultaneous localization and tracking (SLAT) method of the UAV involved network via non-parametric belief propagation (NBP) algorithm, where the UAVs and vehicles are located and tracked jointly. Numerical simulations shed light on the benefits of our proposed method in terms of localization and tracking accuracy.

Unmanned aerial vehicles (UAVs) are attractive in some Internet of Things (IoT) applications, due to their flexible deployment and extended coverage. In this paper, we consider an UAV-assisted network where the UAV flies between the resource-limited sensor nodes (SNs) and collects their status updates. The UAV trajectory and SN scheduling are jointly optimized to minimize the Age of Synchronization (AoS). In contrast to the conventional Age of Information (AoI), AoS takes into account both the freshness and the content of the information, which makes AoS a more suitable design criterion for information collection in an energy-constrained wireless network. Since the formulated problem is challenging to solve due to its non convexity, we reformulate the problem as a Markov decision process (MDP) and propose a deep reinforcement learning (DRL) algorithm to obtain the optimal solution with various action and state spaces. Our simulation results show the fast convergence rate of the proposed DRL algorithm and demonstrate that our proposed scheme can improve the performance of the UAV-assisted network compared to AoI-based schemes.

In recent years, unmanned aerial vehicles (UAV) have been widely adopted to assist sensing, localization and communication for ground devices (GDs), especially for emergency applications. Since the location of GDs cannot be perfect known to UAVs in such cases, we present a novel UAV aided sensing, localization and communication integrated system. We first introduce the idea of age of information (AoI) to evaluate the timeliness of information from GDs, which is essentially determined by the time of UAV movement, localization operation and data transmission. In particular, we focus on the time-critical network scheduling policies, which could be formulated as a joint UAV trajectory, UAV three-dimensional location and bandwidth allocation optimization problem. This Non-convex problem can be decoupled into two subproblems, which could be solved by specified low complexity algorithms. Simulation results are provided, and can verify the effectiveness in timeliness of our proposed scheduling strategies.

Recent investigations show that, the traffic throughput and safety could be effectively improved when vehicles follow coordination instructions from a centralized coordination center at road intersections. However, due to the inevitable noise of the vehicle control, sensing, and limited channel resources, the coordination instructions need to be “updated” periodically. Aiming at this issue, we define a “Yaw Risk” based re-planning strategy, which consists of a multi-vehicle re-planning selection and a multi-channel allocation scheme, to minimize the uncertainty of the entire coordination system. Variable safety redundancies of the collision-free tunnel are adopted to guarantee the tradeoff between safety and traffic throughput. Numerical results are provided and verify our analysis.

High-precision location information formulates the basis of the modern Internet of Things (IoT). However, since the navigation signals from the global navigation satellite systems (GNSSs) are frequently attenuated or blocked in urban areas, reliable and high accuracy positioning alternatives are thus required for ground devices (GDs). Due to the advantages of their flexible deployment and extensive coverage, unmanned aerial vehicles (UAVs) show significant potential in this ground localization enhancement system. In this article, we propose a UAV aided positioning (UAP) system for GDs, where the UAVs provide valuable flying Line of Sight (LoS) observations. Specifically, we first give the fundamental limits of the proposed UAP system in terms of the Cramer–Rao low bound (CRLB), where the UAVs are treated as “agents” with unknown positions instead of anchors. Then, we formulate a general UAP method using the nonparametric belief propagation (NBP)-based probabilistic framework, to jointly positioning UAVs and GDs simultaneously. Moreover, a two-step clustering-based solution is given to tackle the data association challenge in the multi-UAV scenarios. We also show that proper data feedback could achieve additional performance advantages without any extra measurements. The optimal multi-UAV deployment strategy is then proposed, by which the potential of the UAP system could be fully characterized. Last but not least, we verify our solutions via numerical simulations and practical experiments, which provide meaningful insights and performance evaluations to the system design and implementations.

High accuracy and seamless position information formulates the basis of many modern wireless applications, such as the Internet of Things (IoT) and intelligent transportation systems (ITSs). In this article, aiming at the ground user equipment (UE) those in the “blind spots,” where only limited navigation signals are provided, the temporary aerial-aided “anchors” such as the unmanned aerial vehicles (UAVs) are introduced as alternating solutions. We first give the general fundamental limits of the three-dimensional space–air–ground-integrated localization networks (SAGILNs) using both time and angle measurements. Unlike most existing investigations, we treat aerial nodes as “agents” whose positions are not known beforehand. We then try to formulate an efficient scheduling strategy, where proper networkbehaviors, including the resource optimization and UAV deployment, are provided. We find that the proposed scheduling problems could be formulated as standard semidefinite programming (SDP) problems and solved by off-the-shelf solvers. Numerical results are provided to validate our analysis. The proposed methods and analyses provide meaningful insights for performance benchmarks for the implementation of SAGILN.

In typical autonomous driving, a lane-level (submeter) accuracy and ubiquitous coverage is required. Since the signals from the widely adopted Global Navigation Satellite Systems (GNSS) are frequently attenuated or blocked in urban areas, reliable and high accuracy positioning alternatives are thus required. In this paper, we propose an unmanned aerial vehicle (UAV) aided vehicle positioning framework, combined with the general non-parametric belief propagation (NBP) method, to improve the positioning accuracy of vehicles at blind spots. Aiming at the data association issue during the UAV detection, a cluster-based two-step joint probabilistic data association (JPDA) method is adopted. Furthermore, we find that in the multi-UAV scenarios, proper messages feedback of vehicles can effectively improve the data association, then further enhance the positioning accuracy. Numerical results are provided, to validate our analysis, and show significant performance advantages.

Seamless high precision positioning formulates the basis for intelligent driving. Due to the frequent occurrence of blind spots for vehicles in urban areas, we try to introduce the unmanned aerial vehicle for ground vehicle localization assistance in this paper. We first formulate the general scheduling framework of the UAV-aided vehicle localization system. Since the proper deployment of UAVs is of great importance to the localization performance. We then give a resource allocation based UAV deployment method. High accuracy iterative algorithms are provided to solve the essential non-convex problem. Simulation results show the convergence of the proposed method in all investigated cases. Furthermore, obvious performance advantages can be achieved.