Fingerprint-based indoor localization is a critical enabling technology for Internet of Things (IoT) applications, where the primary challenges stem from complex environmental variability and prohibitive costs of data collection and labeling. This paper introduces a cross-domain multi-modal indoor localization framework that effectively combines visual and WiFi signals using few-shot learning techniques, achieving improved localization performance with minimal training data. We derive an upper bound on the generalized transfer localization error. Based on this bound, our learning-based approach applies feature-level knowledge distillation from pre-trained localization models. This process systematically calibrates discrepancies in feature distributions between the source and target environments. As a result, the proposed method significantly reduces the dependence on large labeled datasets. Experimental results demonstrate that our proposed method achieves substantial improvements over state-of-the-art localization models, with a mean localization error of 0.247 meters across diverse indoor environments, while requiring substantially fewer labeled samples in the target domain.

Person re-identification (ReID), a crucial technology for intelligent surveillance in Internet of Things (IoT) systems, aims to search for the target person among the non-overlapping surveillance cameras. Video-based clothes-changing person re-identification (VCC-ReID) has become essential due to the rich information in videos and its broad applications. Because clothes are attached to the human body, the clothes and pedestrian features are highly coupled when extracting features, making VCC-ReID challenging. To solve this challenge, we propose a Multi-Task Disentanglement Framework guided by Pedestrian Attributes (MTDF-PAttr), whose core is the cross-domain attribute distillation decoupling mechanism. Pedestrian attribute recognition (PAR) is used as an auxiliary task in MTDF-PAttr to guide feature decoupling, thereby enhancing the main task, VCC-ReID’s performance. Since the existing VCC-ReID dataset lacks PAR annotations, we employ knowledge distillation to train the auxiliary task, where the teacher network is a pre-trained video-based PAR network. To make the PAR teacher network have better accuracy, stronger generalization, and can identify more attributes, we propose a Multi-Dataset Fusion Framework for Pedestrian Attribute Recognition (MDFF-PAttr), whose core is the multi-teacher collaborative self-distillation mechanism. MDFF-PAttr can simultaneously use multiple datasets for training and provide a powerful teacher model for MTDF-PAttr to distill its auxiliary task. Experimental results demonstrate that MTDF-PAttr can achieve state-of-the-art performance in the VCC-ReID task, providing an effective method for intelligent surveillance systems in the IoT. Additionally, MDFF-PAttr can effectively enhance the accuracy and generalization of the PAR network.

Accurate and robust localization is a critical enabler for emerging 5G and 6G applications, including autonomous driving, extended reality (XR), and smart manufacturing. While data-driven approaches have shown promise, most existing models require large amounts of labeled data and struggle to generalize across deployment scenarios and wireless configurations. To address these limitations, we propose a foundation-model-based solution tailored for wireless localization. We first analyze how different self-supervised learning (SSL) tasks acquire general-purpose and task-specific semantic features based on information bottleneck (IB) theory. Building on this foundation, we design a pretraining methodology for the proposed Large Wireless Localization Model (LWLM). Specifically, we propose an SSL framework that jointly optimizes three complementary objectives: (i) spatial-frequency masked channel modeling (SF-MCM), (ii) domain-transformation invariance (DTI), and (iii) position-invariant contrastive learning (PICL). These objectives jointly capture the underlying semantics of wireless channel from multiple perspectives. We further design lightweight decoders for key downstream tasks, including time-of-arrival (ToA) estimation, angle-of-arrival (AoA) estimation, single base station (BS) localization, and multiple BS localization. Comprehensive experimental results confirm that LWLM consistently surpasses both model-based and supervised learning baselines across all localization tasks. In particular, LWLM achieves 26.0%--87.5% improvement over transformer models without pretraining, and exhibits strong generalization under label-limited fine-tuning and unseen BS configurations, confirming its potential as a foundation model for wireless localization.

Accurate mobile device localization is critical for emerging 5G/6G applications such as autonomous vehicles and augmented reality. In this paper, we propose a unified localization method that integrates model-based and machine learning (ML)-based methods to reap their respective advantages by exploiting available map information. In order to avoid supervised learning, we generate training labels automatically via optimal transport (OT) by fusing geometric estimates with building layouts. Ray-tracing based simulations are carried out to demonstrate that the proposed method significantly improves positioning accuracy for both line-of-sight (LoS) users (compared to ML-based methods) and non-line-of-sight (NLoS) users (compared to model-based methods). Remarkably, the unified method is able to achieve competitive overall performance with the fully-supervised fingerprinting, while eliminating the need for cumbersome labeled data measurement and collection.

Multi-access edge computing (MEC) emerges as an effective computational paradigm. It meets the low latency and low energy consumption requirements of users by enabling user terminals (UE) to offload their computationally intensive applications to nearby access points (AP). However, the integration of Virtual Reality (VR) applications within MEC environments remains underexplored, primarily due to their structural complexity and high computational requirements for offloading and scheduling. To address these challenges, we use Directed Acyclic Graph (DAG) to model VR applications and 6G-oriented Rate-Splitting Multiple Access (RSMA) to enhance offloading and scheduling processes of VR applications. The objective is to jointly optimize the offloading strategy, transmit power, RSMA decoding order, and UE application scheduling to minimize the total system cost. Recognizing the limitations of traditional learning-based methods, which struggle with convergence in multi-user scenarios, we decompose the optimization problem into two subproblems: application offloading and application scheduling. We then propose the PPOCO algorithm, which integrates reinforcement learning with convex optimization to effectively solve these subproblems independently. Experimental results demonstrate that our proposed method consistently out-performs baseline algorithms in reducing the total system cost across various MEC network configurations.

Immersive communication, including emerging augmented reality, virtual reality, and holographic telepresence, has been identified as a key service for enabling next-generation wireless applications. To align with legacy wireless applications, such as enhanced mobile broadband or ultra-reliable low-latency communication, network slicing has been widely adopted. However, attempting to statistically isolate the above types of wireless applications through different network slices may lead to throughput degradation and increased queue backlog. To address these challenges, we establish a unified QoS-aware framework that supports immersive communication and legacy wireless applications simultaneously. Based on the Lyapunov drift theorem, we transform the original long-term throughput maximization problem into an equivalent short-term throughput maximization weighted by virtual queue length. Moreover, to cope with the challenges introduced by the interaction between large-timescale network slicing and short-timescale resource allocation, we propose an adaptive adversarial slicing (Ad2S) scheme for networks with invarying channel statistics. To track the network channel variations, we also propose a measurement extrapolation-Kalman filter (ME-KF)-based method and refine our scheme into Ad2S-non-stationary refinement (Ad2S-NR). Through extended numerical examples, we demonstrate that our proposed schemes achieve 3.86 Mbps throughput improvement and 63.96% latency reduction with 24.36% convergence time reduction. Within our framework, the trade-off between total throughput and user service experience can be achieved by tuning systematic parameters.

Extended reality (XR), blending virtual and real worlds, is a key application of future networks. While AI advancements enhance XR capabilities, they also impose significant computational and energy challenges on lightweight XR devices. In this article, we developed a distributed queue model for multitask deep neural network inference, addressing issues of resource competition and queue coupling. In response to the challenges posed by the high energy consumption and limited resources of XR devices, we designed a dual time-scale joint optimization strategy for model partitioning and resource allocation, formulated as a bi-level optimization problem. This strategy aims to minimize the total energy consumption of XR devices while ensuring queue stability and adhering to computational and communication resource constraints. To tackle this problem, we devised a Lyapunov-guided proximal policy optimization algorithm, named LyaPPO. Through numerical results, we show that our LyaPPO algorithm outperforms the baseline algorithms. Specifically, under different maximum local computational capacities, the proposed algorithm decreases 24.29%–56.62% energy compared to the suboptimal baselines.

This paper presents a novel two-stage deep reinforcement learning (DRL) algorithm built on a Transformer Encoder-based Deep Deterministic Policy Gradient (TEDDPG) framework, named TS-TEDDPG, which jointly optimizes the learning latency, energy consumption and model accuracy of Asynchronous Federated Learning (AFL) systems with prescribed security. The CPU configuration of local training and the transmit power of model uploading are learnt by the TEDDPG in the first stage. A linear programming-based device scheduling and cooperative jamming strategy is designed to efficiently optimize the rest of the decisions in the second stage and evaluates the immediate reward to train the TEDDPG. Experimental results based on a CNN model and the MNIST dataset demonstrate that the proposed TS-TEDDPG can reduce the training latency and energy consumption by 68.6% compared to its benchmarks, when the required test accuracy is 0.9.

Extended reality (XR) is one of the most important applications of beyond 5G and 6G networks. Real-time XR video transmission presents challenges in terms of data rate and delay. In particular, the frame-by-frame transmission mode of XR video makes real-time XR video very sensitive to dynamic network environments. To improve the users’ quality of experience (QoE), we design a cross-layer transmission framework for real-time XR video. The proposed framework allows the simple information exchange between the base station (BS) and the XR server, which assists in adaptive bitrate and wireless resource scheduling. We utilize the cross-layer information to formulate the problem of maximizing user QoE by finding the optimal scheduling and bitrate adjustment strategies. To address the issue of mismatched time scales between two strategies, we decouple the original problem and solve them individually using a multi-agent-based approach. Specifically, we propose the multi-step Deep Q-network (MS-DQN) algorithm to obtain a frame-priority-based wireless resource scheduling strategy and then propose the Transformer-based Proximal Policy Optimization (TPPO) algorithm for video bitrate adaptation. The experimental results show that the TPPO+MS-DQN algorithm proposed in this study can improve the QoE by 3.6% to 37.8%. More specifically, the proposed MS-DQN algorithm enhances the transmission quality by 49.9%-80.2%.

The advent of online video services has resulted in a remarkable surge in Internet traffic, prompting the need for mobile edge computing (MEC) as a crucial element in augmenting the quality of adaptive streaming media services amidst the time-varying wireless channels. MEC reduces network backhaul traffic by providing video transcoding and adaptive streaming services closer to users. Nonetheless, the process of video transcoding introduces additional latency and energy consumption. In order to effectively tackle this challenge and uphold the optimal quality of experience (QoE), we propose the Joint Bitrate Transcoding and Parallel Cooperative Transmission (JBTPCT) model, which operates at the edge of mobile networks and handles multiple video chunks simultaneously. Within the JBTPCT model, the Asynchronous Advantage Actor-Critic (A3C) algorithm framework is employed to jointly account for radio access network conditions and MEC resources, leveraging a parallel execution strategy for transmission and transcoding. This integrated approach aims to minimize both latency and energy consumption while enhancing the QoE of video streaming. We evaluate the average QoE of JBTPCT in different network scenarios, and the experimental results demonstrate that JBTPCT consistently achieves higher average QoE compared to competing algorithms.

Cooperative inference in Mobile Edge Computing (MEC), achieved by deploying partitioned Deep Neural Network (DNN) models between resource-constrained user equipments (UEs) and edge servers (ESs), has emerged as a promising paradigm. Firstly, we consider scenarios of continuous Artificial Intelligence (AI) task arrivals, like the object detection for video streams, and utilize a serial queuing model for the accurate evaluation of End-to-End (E2E) delay in cooperative edge inference. Secondly, to enhance the long-term performance of inference systems, we formulate a multi-slot stochastic E2E delay optimization problem that jointly considers model partitioning and multi-dimensional resource allocation. Finally, to solve this problem, we introduce a Lyapunov-guided Multi-Dimensional Optimization algorithm (LyMDO) that decouples the original problem into per-slot deterministic problems, where Deep Reinforcement Learning (DRL) and convex optimization are used for joint optimization of partitioning decisions and complementary resource allocation. Simulation results show that our approach effectively improves E2E delay while balancing long-term resource constraints.

The high computational complexity and high energy consumption of artificial intelligence (AI) algorithms hinder their application in augmented reality (AR) systems. This paper considers the scene of completing video-based AI inference tasks in the mobile edge computing (MEC) system. We use multiply-and-accumulate operations (MACs) for problem analysis and optimize delay and energy consumption under accuracy constraints. To solve this problem, we first assume that offloading policy is known and decouple the problem into two subproblems. After solving these two subproblems, we propose an iterative-based scheduling algorithm to obtain the optimal offloading policy. We also experimentally discuss the relationship between delay, energy consumption, and inference accuracy.

Network slicing is a critical driver for guaranteeing the diverse service level agreements (SLA) in 5G and future networks. Recently, deep reinforcement learning (DRL) has been widely utmzed for resource allocation in network slicing. However, existing related works do not consider the performance loss associated with the initial exploration phase of DRL. This paper proposes a new performance-guaranteed slicing strategy with a soft and hard hybrid slicing setting. Mainly, a common slice setting is applied to guarantee slices’ SLA when training the neural network. Moreover, the resource of the common slice tends to precisely redistribute to slices with the training of DRL until it converges. Furthermore, experiment results confirm the effectiveness of our proposed slicing framework: the slices’ SLA of the training phase can be guaranteed, and the proposed algorithm can achieve the near-optimal performance in terms of the SLA satisfaction ratio, isolation degree and spectrum efficiency after convergence.

With the rapid development of indoor location-based services (LBSs), the demand for accurate localization keeps growing as well. To meet this demand, we propose an indoor localization algorithm based on graph convolutional network (GCN). We first model access points (APs) and the relationships between them as a graph, and utilize received signal strength indication (RSSI) to make up fingerprints. Then the graph and the fingerprint will be put into GCN for feature extraction, and get classification by multilayer perceptron (MLP). In the end, experiments are performed under a 2D scenario and 3D scenario with floor prediction. In the 2D scenario, the mean distance error of GCN-based method is 11m, which improves by 7m and 13m compare with DNN-based and CNN-based schemes respectively. In the 3D scenario, the accuracy of predicting buildings and floors are up to 99.73% and 93.43% respectively. Moreover, in the case of predicting floors and buildings correctly, the mean distance error is 13m, which outperforms DNN-based and CNN-based schemes, whose mean distance errors are 34m and 26m respectively.

Ideally, the throughput of a Multipath TCP (MPTCP) connection should be as high as that of multiple disjoint single-path TCP flows. In reality, the throughput of MPTCP is far lower than expected. This is fundamentally caused by the fact that a sub flow with high delay and loss affects the performance of other sub flows, and thus becomes the bottleneck of the MPTCP connection and significantly degrades the aggregate good put. To tackle this problem, we propose Fountain code-based Multipath TCP (FMTCP), which effectively mitigates the negative impact of the heterogeneity of different paths. FMTCP takes advantage of the random nature of the fountain code to flexibly transmit encoded symbols from the same or different data blocks over different sub flows. Moreover, we design a data allocation algorithm based on the expected packet arriving time and decoding demand to coordinate the transmissions of different sub flows. Quantitative analyses are provided to show the benefit of FMTCP. We also evaluate the performance of FMTCP through ns-2 simulations and demonstrate that FMTCP can outperform IETF-MPTCP, a typical MPTCP approach, when the paths have diverse loss and delay in terms of higher total good put, lower delay and jitter. In addition, FMTCP achieves much more stable performance under abrupt changes of path quality.