Intelligent reflecting surfaces (IRSs) mounted on maneuverable aerial platforms to form aerial IRS (AIRS) relays represent a novel paradigm for large-scale downlink transmission in smart cities. However, the challenge of multivariate dynamic coupling hinders most existing studies due to high computational complexity and limited scalability. To address these issues, this paper proposes a soft-attention-actor-critic (SAAC) optimization framework that efficiently decomposes the joint optimization of multi-AIRS deployment, passive beamforming, and active beamforming at the base station into two sequential subproblems. The objective is to maximize average downlink spectral efficiency and service fairness, while minimizing deployment energy consumption. In the first stage, a conservative lower bound of spectral efficiency is formulated to guide multiple AIRSs toward near-optimal deployment positions. In the second stage, refined optimization is performed for both passive and active beamforming matrices. Furthermore, multi-head attention modules are incorporated into the critic and actor networks in each phase, enabling AIRS to adaptively attend to the observations and actions of other agents, and enhancing the ability to handle high-dimensional observation-action spaces. Extensive simulation results validate that the proposed SAAC framework consistently outperforms mainstream deep reinforcement learning baselines across diverse network conditions, highlighting its superior performance and scalability.

Mobile crowdsensing (MCS) enables data collection by leveraging the sensing capabilities of distributed mobile devices (MDs). However, its performance is often constrained by limited coverage and connectivity of ground-based sensors. To overcome these limitations, this paper presents an efficient data collection framework in unmanned aerial vehicle (UAV)-assisted MCS systems. The proposed framework leverages the aerial mobility of UAVs to enhance the spatial coverage of MCS. By optimizing flight trajectories of UAVs, we aim to balance the age of information (AoI) and energy consumption. Specifically, this work considers the mobility of MDs and introduces a communication-enhanced trajectory optimization (CETO) algorithm to improve UAV coordination and adaptability in dynamic environments. Simulation results demonstrate that the proposed algorithm significantly outperforms other mainstream baseline methods employing deep reinforcement learning (DRL).

This paper investigates the joint design of the transmit beamforming matrix and the intelligent reflecting surface (IRS) phase shift matrix in a multi-user, multiple-input singleoutput (MU-MISO) system integrated with an aerial IRS. The proposed approach leverages the high mobility and flexibility of unmanned aerial vehicles (UAVs) to enhance the probability of a line-of-sight (LoS) link, thereby maximizing the sum spectral efficiency of the user equipments. Specifically, to adapt to the time-varying nature of real-world communication environments, we propose a twin delayed deep deterministic policy gradient (TD3) algorithm incorporating self-attention mechanisms for the joint optimization of beamforming and phase shift strategies. Furthermore, the batch normalization technique is employed to improve the algorithm's capability to process extensive state and action spaces, thereby accelerating convergence. Simulation results demonstrate that the proposed algorithm outperforms other mainstream deep reinforcement learning (DRL)-based baseline methods.

In the realm of the Internet of Things (IoT), computation offloading confronts challenges in remote areas due to scarce general-purpose edge/cloud infrastructure and insufficient terrestrial network coverage. To address this, we introduce a novel space-air-ground integrated network (SAGIN) computing architecture, designed for the efficient offloading of computationintensive applications. Within this architecture, unmanned aerial vehicles (UAVs) conduct edge computing near users, while satellites act as a bridge to cloud computing resources. Given the limitations of UAVs in terms of battery capacity and dynamic network topology, their deployment strategy is crucial for maintaining service quality. Due to the impracticality of collecting global user information for centralized control of UAVs, we have conducted research on the adaptive deployment of UAVs under the condition that they rely solely on local observations. We propose a multi-agent softmax deep double deterministic policy gradient (MASD3) algorithm and comprehensively consider maximizing the uplink transmission rate of terrestrial IoT devices and reducing the energy consumption of UAVs during flight and communication in the optimization objective. Simulation results demonstrate that our proposed solution outperforms existing state-of-the-art baselines.

Transformer-based models have demonstrated outstanding performance in video understanding tasks due to their capacity to capture long-range dependencies. However, their high computational cost, along with the massive volume of streaming video data, presents significant challenges for real-time deployment on resource-constrained edge devices integrated into internet of things (IoT) systems. Existing approaches typically eliminate spatial or temporal redundancy in isolation, failing to fully exploit the inherent spatio-temporal similarity in video data. To address this limitation, we propose CST-ViT, a cascaded spatio-temporal redundancy elimination framework that jointly reduces dynamic temporal and intra-frame spatial redundancy. CST-ViT incorporates three gating modules: the direct temporal gate for matching unchanged backgrounds, the offset temporal gate for capturing motion-related changes, and the spatial gate for intra-frame similarity matching. Together with a spatiotemporal caching and token reuse mechanism, CST-ViT enables efficient token filtering and computation reuse. Experimental results show that CST-ViT reduces computation by 55.88% with no loss in accuracy, and achieves up to a 74.75% reduction in computation with less than 1% accuracy degradation, outperforming state-of-the-art methods in terms of accuracy–efficiency trade-off for video transformers.

This paper presents an intelligent reflecting surface (IRS)-assisted data harvesting scheme for unmanned aerial vehicle (UAV) networks. This scheme leverages the high maneuverability of the UAV and the channel gain enhancement from the IRS. By jointly optimizing the UAV trajectory and the IRS phase shift, we aim to minimize the completion time of data harvesting missions. Specifically, we devise a softmax operator applicable to deterministic policy gradients and propose a softmax deep double deterministic policy gradients (SD3) method to facilitate the design of three-dimensional trajectory for UAV. In addition, we propose a practical coherent combining (CC) strategy for IRS phase control. Simulation results demonstrate that the proposed SD3-CC algorithm surpasses other mainstream baseline methods relying on deep reinforcement learning (DRL).

Most congestion control mechanisms perform well in specific network environments, but none can consistently deliver good performance across all scenarios. Recently proposed frameworks based on reinforcement learning can flexibly select congestion control algorithms to adapt to dynamic changes in network conditions. However, frequently altering the congestion control mechanisms during relatively stable periods of the network actually leads to instability and unnecessary computational overhead. In this paper, we propose a hierarchical adaptive congestion control algorithm (HACC) to be resilient to the varying network. HACC dynamically selects the appropriate congestion control mechanism only when the current congestion control algorithm is not suitable for the current network state, rather than changing the congestion control scheme every training cycle to ensure network stability. The simulation results show that under different realistic workloads, HACC significantly reduces the computational overhead and improves throughput. Specifically, HACC reduces average overhead by 31% and improves throughput by up to 47%, 35%, 23%, and 15% compared to Cubic, Reno, BBR, and Antelope, respectively.

Modern data centers are hosting a great number of various applications (e.g. MapReduce and web search) that require a high fan-in data communication, which easily causes serious packet losses and timeouts, substantially degrading the application performance. To address this issue, various host-based and switch-based transport protocols are proposed to eliminate timeout and improve the user experience. Unfortunately, although existing transport protocols can effectively eliminate the timeout, they inevitably result in persistent queueing backlog and degrade the network performance, especially delay-sensitive short flows. To this end, we propose a general scheme with ultra-low latency called UL2to address the above problem. Concretely, the sender periodically estimates the queueing delay of each packet on the transmission path and senses the degree of congestion based on its measured result. Then the sender timely yet cautiously executes a pausing transmission operation based on measured queueing delay, guaranteeing fast elimination of queue delays and high link utilization. Our evaluation indicates that UL2can effectively eliminate queue backlog and reduce the queueing delay by more than 90%. Moreover, UL2enhances the performance of state-of-the-art transport protocols in terms of flow completion times by up to 44.98%.

The collaboration between unmanned aerial vehicles (UAVs) and intelligent reflecting surfaces (IRSs) presents an innovative approach for delay-tolerant data harvesting in distributed Internet of Things (IoT) networks. However, existing research mostly overlooks the dynamic changes in communication links caused by the real-time UAV movement and the realistic geographical features. In this paper, we address these challenges by considering a practical three-dimensional (3D) urban scenario with a centralized IRS. Our aim is to minimize the completion time of data harvesting missions by jointly optimizing the 3D trajectory of the UAV and the phase shift of the IRS. Specifically, the formulated problem is decoupled into two subproblems. First, for the 3D continuous trajectory design, we propose a robust memory-based softmax deep double deterministic policy gradients (MSD3) approach, which enables the UAV to adaptively collect delay-tolerant data from randomly distributed ground devices starting from any arbitrary point. Second, we present a comprehensive theoretical analysis for the continuous IRS phase control, which provides a practical and intuitive numerical solution. Simulation results demonstrate that the proposed MSD3-IRS algorithm outperforms other mainstream baselines based on deep reinforcement learning.

With the development of Internet of Things (IoT), wireless communication networks and Artificial Intelligence (AI), more and more real-time applications such as online games and autonomous driving have emerged. However, due to limited computing power and battery capacity, it has become increasingly difficult for local user devices to take on the full range of computing tasks under tight timing constraints. The emerging Mobile Edge Computing (MEC) technology is widely considered to be an important technology for achieving ultra-low latency. However, most of the existing work is focused on non-splittable computation tasks. In fact, data partitioning-oriented applications can be split into multiple subtasks for parallel processing. In this paper, we study the partial computation offloading of multiple detachable tasks in MEC networks, focusing on minimizing the total user device latency in the multi-MEC multi-user scenarios. Considering the dynamic partitioning of tasks, we adopt the barrel theory to construct a linear system of equations to find the optimal solutions and propose an approach for distributed computation offloading based on numerical methods. The simulation results show that the proposed algorithm can reduce the average user device latency by 31 % compared with the binary offloading method.