Search

In vehicular edge computing networks, the real-time, on-demand scheduling of scarce network resources for environmental perception, task offloading, computation, and feedback is vital. However, these coupled processes make resource allocation challenging. Moreover, existing real-time channel measurement techniques in complex vehicular topologies present load, accuracy, and customization difficulties. To address these issues, this paper proposes an on-demand environmental perception and resource allocation strategy. Specifically, with the introduction of a channel knowledge base, we first analyze the coupling relationship between environmental perception, communication, and computation. A model is then proposed for task offloading to schedule the granularity of environment perception, communication resources, and computational resources dynamically. Subsequently, the resource allocation problem is formulated as an optimization problem, aiming to minimize system processing delay and maximize resource utilization while ensuring perception accuracy. To address this, a two-phase optimization-assisted deep reinforcement learning (DRL) algorithm is proposed. The initial phase uses convex optimization to approximate a solution. The second phase proposes a DRL-based algorithm to intelligently schedule dynamic network resources, with the first phase’s solution guiding the initial exploration space to enhance DRL training efficiency. Extensive simulation experiments verify the effectiveness of our proposal.

Air-ground Integrated Mobility (AIM), as a disruptive mode of travel, has the tremendous potential to alleviate ground traffic congestion issues substantially. However, the primary challenge in achieving this leapfrog development lies in ensuring driving safety. Receiving collision warnings in time within a limited distance can significantly reduce collision risks, which is crucial for ensuring driving safety in AIM. However, due to challenges in aerial network coverage, ensuring the communication quality of aerial Personal Aerial Vehicles (PAVs) remains difficult, thereby affecting the effective transmission of messages. Furthermore, the integration of ground Connected and Automated Vehicles (CAVs) with aerial PAVs in AIM results in significant differences in user resource requirements. Given the complexity of the AIM environment and the high mobility of PAVs, it is challenging to rapidly and accurately capture user communication quality. Therefore, addressing the differential resource requirements of users in this environment is particularly challenging. To this end, we propose a personalized resource allocation strategy assisted by a Channel Knowledge Map (CKM) in AIM. This strategy aims to meet the personalized resource requirements of users while maintaining the maximum Perception Response Time (PRT), thereby ensuring driving safety. Specifically, the CKM in AIM is constructed to obtain channel states through environment-aware communication. Next, a 3D collision warning system is designed to analyze rigorously the maximum PRT of vehicles under different motion states in avoiding collisions. On this basis, with the help of CKM, the channel knowledge of the user’s location is obtained to quantify the communication and computing resources required by each user to maintain the maximum PRT. Finally, we establish the PRT-driven resource optimization problem and employ Deep Reinforcement Learning (DRL) to seek the optimal resource allocation strategy. Simulation results indicate that the proposed method effectively enhances safety and resource utilization in AIM under resource constraints and uneven distribution.

Preemptive scheduling efficiently addresses the coexistence of enhanced Mobile Broad Band (eMBB) and Ultra-Reliable Low-Latency Communications (URLLC). While URLLC puncturing influences eMBB performance, further investigation is necessary to study the trade-offs between stability, delay, and efficiency. However, existing studies overlook the imbalance in eMBB/URLLC load distribution and personalized fluctuations in eMBB performance, leading to sub-optimal results. To tackle this, we propose an unmanned aerial vehicle (UAV) relay-assisted eMBB/URLLC multiplexing framework. Specifically, considering the utilization of UAVs for connecting separated next-generation Node Bs (gNBs) and the individual subject experience of services, we first formulate the multiplexing problem as an optimization problem. The objective is to maximize eMBB throughput and minimize personalized fluctuations in eMBB performance and UAV consumption, subject to URLLC constraints. Then, the challenging problem is decomposed into the eMBB problem and the URLLC problem. For the former, we further decompose it into three sub-problems and solve them using optimization methods. For the latter, we propose a deep reinforcement learning-based algorithm to obtain an optimal strategy for relaying and puncturing URLLC into eMBB intelligently. Simulation results demonstrate that our proposals outperform benchmark schemes regarding eMBB throughput, UAV consumption, eMBB performance fluctuation, URLLC satisfaction, and learning efficiency.

Vehicular networks have evolved to a new stage where they integrate sensing, communication, and computing capabilities, giving rise to a multitude of vehicular applications that cater to contemporary demands. These applications are characterized by a high degree of integration, coupled functionality between sensing, communication, and computing (SCC), and the need for timely scheduling. Most studies on the integration of sensing, communication, and computing (ISCC) for vehicular networks focus on directly matching SCC resources to task demands. However, in the era of ISCC, the interdependence among tasks is critical and therefore cannot be ignored during the task scheduling process. For instance, the computing task can only start after the sensing task is finished. In addition, the SCC resources and task demands fluctuate significantly as time goes by due to the high mobility of vehicular networks. In this paper, we propose a dependency-aware task scheduling strategy for ISCC-based vehicular networks, which takes both task interdependence and high mobility into consideration. With the proposed strategy, the demands of vehicle application tasks on SCC resources are determined after the relationship between the tasks is examined. In addition, the mobility of vehicles is taken into consideration in order to properly match the demands of the sources on different vehicles. Finally, a meta deep reinforcement learning-based task scheduling (MTS) algorithm is used to make the appropriate task scheduling decision. Extensive simulation results indicate that the proposed strategy can effectively reduce dependent task processing delay in dynamic vehicular networks. In addition, the MTS approach ensures that the proposed strategy can quickly adapt to new vehicular network environments.

Air-ground Integrated Mobility (AIM) can effectively alleviate the current urban traffic pressure by expanding transportation resources in the near-ground field. However, the following problems in AIM need to be addressed urgently: 1) The high mobility of Personal Aerial Vehicles (PAVs) in low-altitude airspace leads to a sharp increase in risk factors; 2) Due to the limited communication distance, antenna direction angle, and frequent handover caused by high-speed movement, the communication quality in the air is unreliable; 3) AIM incorporates vehicles on the ground and PAVs in the air leading to the high variability of user requirements. Confronted with the personalized resource requirements of high-speed mobile PAVs in airspace with unreliable communication quality, traditional resource allocation strategies struggle to guarantee service quality. Therefore, we propose a safety-oriented personalized resource allocation strategy in AIM, which jointly considers the user requirements and resource distribution. Specifically, we first build a three-dimensional (3D) safety distance model by analyzing the motion process of PAVs with the help of a kinematics model. Then, according to the location, speed, and environmental information of the PAVs, the communication and computing resources required by each PAV under the premise of maintaining the optimal safety distance are obtained through the transmission model. Furthermore, the 3D safety distance and resources are jointly optimized, and an on-demand resource allocation algorithm enabled by Deep Reinforcement Learning (DRL) is constructed to provide the resource allocation strategy based on the personalized requirements of the users.

Unmanned aerial vehicle (UAV) relay networks with flexible and controllable characteristics are expected to complement the capacity of the gNB. This paper studies the multiplexing of enhanced Mobile BroadBand (eMBB) and Ultra-Reliable Low-Latency Communications (URLLC) in a multi-UAV relay network, where the strict latency requirement of URLLC can be achieved by the preemptive multiplexing of eMBB resources. However, this may affect eMBB reliability due to the transmission interruptions. Moreover, given the limited energy resources of UAVs, there is an inherent tradeoff among reliability, delay, spectral efficiency, and energy efficiency. To address these challenges, this paper develops a hierarchical UAV-assisted eMBB/URLLC multiplexing scheduling framework. For the eMBB scheduler, we first utilize multiple UAVs to assist the gNB in relaying eMBB traffic and formulate the eMBB resource allocation problem as an optimization problem. Then, we propose a decomposition-relaxation-optimization algorithm to maximize eMBB data rates while considering the personalized fairness of resource allocation and UAV power consumption. For the URLLC scheduler, we further consider the multiplexing of eMBB/URLLC traffic based on the optimization of eMBB resources. To reduce the performance fluctuations of eMBB, we propose a novel cross-slot strategy to schedule URLLC within two time slots rather than one time slot as in existing works. With this strategy, a deep reinforcement learning-based algorithm is proposed to obtain the optimal strategy for the preemption of URLLC on eMBB. Simulation results show that the proposed algorithms outperform the benchmark schemes in terms of convergence rate, eMBB reliability, personalized resource fairness, UAV consumption, and URLLC satisfaction.