Searcharxiv⌕ Search

arXiv subjects

Qiong Wu

Publications and source records attributed to Qiong Wu.

At least 19 recordsLinked to original sources
arXiv2026

AP Association for RHS-Enabled Cell-Free Uplink MIMO in Industrial Indoor UAV Networks

Indoor industrial UAV uplink networks face serious blockage and shadowing from shelves, metal equipment, and production facilities. UAVs are also often clustered and fly along similar straight inspection routes at fixed heights. These features make traditional small-cell deployment less suitable, especially when high reliability, continuous coverage, and good service for weak UAVs are required. Cell-free networks can improve robustness through distributed access points (APs) and UAV?centric communications. Reconfigurable holographic surface (RHS)-enabled APs provide programmable analog receive beams and generate scalar post-RHS observations, which are jointly processed at the CPU for distributed uplink MIMO detection at relatively low hardware cost. Conventional AP association relies on distance, large-scale fading, or post-combining SINR obtained with user-specific digital combiners. Here, however, all UAVs served by a single-feed RHS AP share one amplitude?constrained receive pattern and one scalar AP output. We therefore derive an SINR-like score from this physical output and, under weak inter-AP disturbance correlation, approximate the CPU-side log-det objective by an additive per-AP surrogate, yielding a low-complexity ranking rule. The results show that the nearest AP is not always the best choice, the AP-UAV height difference may have an optimal value, and larger serving clusters bring diminishing returns. Simulations show that the proposed method improves the minimum UAV data rate, average spectral efficiency, fairness, and energy efficiency compared with benchmark schemes.

cs.NI↗
arXiv2026

Active Movable-Element RIS Assisted Vehicular Semantic Communications: Modeling and Optimization

Severe signal blockage and fast-varying channels in vehicular environments pose critical challenges to reliable semantic communication. To address these, this paper proposes a novel Row-Movable Active Reconfigurable Intelligent Surface (RM-A-RIS) assisted vehicular semantic communication system. This architecture uniquely combines active signal amplification with element mobility to compensate for multiplicative fading and reconstruct channel geometry, thereby enhancing spatial diversity. We formulate a joint optimization problem to maximize Semantic Spectral Efficiency (SSE) by coordinating RIS element positions, active reflection coefficients, and semantic symbol length. An efficient Alternating Optimization (AO) algorithm is developed to tackle the coupled non-convexity. Simulation results demonstrate that the proposed scheme substantially outperforms existing benchmarks, achieving up to 132.9%, 9.2%, and 35.2% improvements in Sum-Semantic Spectral Efficiency (Sum-SSE) compared to the passive RIS, fixed-position active RIS, and QPSO baselines, respectively.

cs.NI↗
arXiv2026

Mobile Tracking via Target-Mounted IRS-Assisted ISAC System

This paper proposes a target-mounted intelligent reflecting surface (IRS)-assisted integrated sensing and communication framework for real-time unmanned aerial vehicle (UAV) tracking, addressing challenges such as link blockage and weak radar cross section in the low-altitude economy. By integrating the IRS onto the UAV, the system creates a mobile cooperative target that provides controllable line-of-sight echoes for self-tracking while acting as a mobile relay for ground communication enhancement. We establish a comprehensive three dimensions state evolution model for the maneuvering UAV. Based on this model, an extended Kalman filter is immediately implemented to achieve real time tracking of the moving UAV. To characterize the fundamental theoretical limits of this recursive estimation process, we derive the analytical posterior Cramer Rao bound and a closed form expression for the elliptical tradeoff performance bound to quantify the relationship between sensing precision and communication throughput. To ensure millisecond level responsiveness, we develop a low complexity joint beamforming design. By utilizing the analytical mapping between tracking and communication requirements, the proposed scheme yields closed form solutions for beamforming vectors, effectively bypassing the time consuming numerical iterations of conventional methods. Numerical simulations demonstrate that the proposed framework significantly outperforms traditional fixed-deployment benchmarks across complex maneuvering trajectories, achieving centimeter-level accuracy while substantially reducing transmit power and processing latency.

eess.SP↗
arXiv2026

Convergence-Latency-Aware Adaptive Modulation and Resource Allocation in RIS-Assisted Wireless Federated Learning

Federated learning (FL) over wireless networks suffers from significant training latency and degraded convergence due to unreliable wireless transmission, especially under blocked propagation environments. Although reconfigurable intelligent surfaces (RISs) can improve communication reliability, existing wireless FL studies rarely characterize the trade-off between learning convergence and communication delay under modulation-dependent transmission errors. In this paper, we consider a wireless FL system operating under RIS-assisted blocked-link propagation scenarios, and focus on adaptive modulation and sub-channel allocation for convergence-latency aware communication design. By characterizing the effect of symbol errors on uploaded local gradients, we derive a convergence-related upper bound that reveals the impact of symbol error rate (SER) on FL loss decay. Based on this result, we formulate a joint convergence-latency optimization problem, which is cast as a mixed-integer nonlinear programming (MINLP) problem, and solve it using a low-complexity hybrid alternating optimization framework. Extensive experiments on MNIST, CIFAR-10, and Speech Commands show that the proposed scheme consistently achieves faster convergence and higher test accuracy than existing adaptive communication schemes, especially in complex tasks and challenging wireless scenarios.

cs.LG↗
arXiv2026

Transmissive RIS Transceiver-Empowered ISAC Systems: Energy Efficiency Optimization for Perfect and Imperfect CSI

In this paper, a novel transmissive reconfigurable intelligent surface (TRIS) transceiver is employed to enable an integrated sensing and communication (ISAC) system supporting both communication and sensing. Under both perfect and imperfect channel state information (CSI), we study the transmit beamforming design for the TRIS transceiver to maximize the system energy efficiency (EE), subject to per-user minimum-rate guarantees, a minimum beampattern gain toward the sensing target, and per-antenna power constraints. The corresponding EE maximization problems are challenging to solve due to the fractional objective and non-convex constraints. In particular, under imperfect CSI, the resulting semi-infinite constraints further complicate the problem. For the perfect CSI case, we first apply the fractional programming (FP) methodology to obtain more tractable reformulations of the rate functions, and then propose an iterative algorithm based on the majorization-minimization (MM) framework. For the imperfect CSI case, we utilize the S-Procedure to transform the semi-infinite inequality constraints into linear matrix inequalities (LMIs), and further develop an efficient MM-based algorithm with the aid of slack variables. Numerical results demonstrate the convergence and effectiveness of the proposed algorithms and validate the EE gains of the TRIS transceiver-enabled ISAC system.

eess.SP↗
arXiv2026

Transmit Beamforming Design for Integrated Sensing and Communication Using Transmissive RIS Transceiver

Integrated sensing and communication (ISAC) is a key technology for future wireless networks, calling for hardware-efficient architectures to jointly support communication and sensing. In this paper, a transmissive reconfigurable intelligent surface (TRIS) transceiver is leveraged to enable an ISAC system. Under the considered system model, we investigate transmit beamforming design for the TRIS transceiver to maximize the sum-rate/beampattern gain, subject to the predefined sensing beampattern gain/communication rate thresholds and the per-unit power constraints of the TRIS transceiver. Since the objective functions and constraints are non-convex, the above two optimization problems are highly challenging. To resolve the difficult optimization problems, we combine the fractional programming (FP) method and the majorization-minimization (MM) framework to develop second-order cone programming (SOCP)-based solutions. Since the per-element power constraints introduce a large number of constraints, this increases the complexity of solving the optimization problems. By splitting the coupling constraints and applying the alternating direction method of multipliers (ADMM) framework, we propose two analytic-based algorithms for efficiently updating the beamformer configurations in the sum-rate and beampattern gain maximization problems, respectively. Simulation results demonstrate the convergence and effectiveness of the proposed algorithms, and show that the low-complexity algorithms achieve performance close to the SOCP-based benchmarks with substantially reduced computational complexity.

eess.SP↗
arXiv2026

Ultrafast giant enhancement of second harmonic generation in a strongly correlated cobaltite

In quantum materials, nonlinear optical responses are highly sensitive to electronic structure and many-body interactions. Probing and manipulating such nonlinear processes is a complex and subtle endeavor, yet it offers deep insights into emerging physics and functionalities. Here, we report an anomalous ultrafast enhancement of second harmonic generation (SHG) in a strongly correlated cobaltite YbBaCo$_4$O$_7$. Above-bandgap femtosecond pumping increases SHG intensity by up to 60 % within 200 fs, with the enhancement persisting for tens of picoseconds. The enhancement is strongly anisotropic, with substantial amplification of in-plane susceptibility tensors, whereas the out-of-plane counterpart shows negligible change. We attribute these anomalies to ultrafast photodoping-induced modulation of the on-site Coulomb repulsion, which dynamically renormalizes the band structure and selectively amplifies specific $χ^{(2)}$ components. These nonlinear dynamics encode rich information about the orbital symmetries and energies of the states involved, opening new avenues for all-optical probing of electronic structure in strongly correlated materials.

cond-mat.str-el↗
arXiv2026

Identifying and Resolving Pitfalls of Knowledge-Based VQA Benchmarks: Auditing, Repairing, and Augmenting

Knowledge-Based Visual Question Answering (KB-VQA) aims to evaluate whether Visual Language Models (VLMs) can retrieve, ground, and reason over external structured knowledge beyond visual evidence. In practice, answer accuracy is widely adopted as the primary evaluation metric, implicitly treating correctness as a proxy for knowledge-grounded reasoning. However, for existing KB-VQA benchmarks, this proxy relies on critical assumptions that are often overlooked and rendered unreliable by benchmark issues: annotated answer must be derivable from the associated knowledge base, question must be well-posed with sufficient constraints, and visual setting must meaningfully require grounded disambiguation. In this work, we show that these assumptions are systematically violated in existing KB-VQA benchmarks. Our audit reveals substantial instances with missing or contradicted answers and underspecified questions that render accuracy a misleading metric. Furthermore, we find that existing datasets rely on visually trivial, single-entity scenes that bypass the need for sophisticated visual-to-knowledge mapping. We demonstrate that even with controlled architectures, these flaws lead to distorted model rankings and overestimations of reasoning capabilities. To address this, we introduce (1) a principled audit-and-repair protocol that restores answer derivability and question clarity, and (2) a controlled multi-entity augmentation protocol that introduces visual ambiguity to challenge initial retrieval and grounded reasoning. Re-evaluation under corrected and augmented settings yields markedly different performance trends. Our findings call for rethinking evaluation protocols and designing more interaction-aware KB-VQA benchmarks that prioritize verifiable reasoning over simple matching.

cs.CL↗
arXiv2026

Towards a Dynamic and Fixed-budget Memory Bank for Efficient Streaming Video Understanding

Currently, streaming video understanding is still a daunting task for existing \emph{multimodal large language models} (MLLMs). Its difficulties not only lie in handling the ever-increasing video frames, but also in the unpredictability of future video content and input instructions. In this paper, we study this task from the perspective of constructing a dynamic but fixed-budget memory bank, and propose a novel and training-free approach termed \emph{\textbf{CausalMem}}. CausalMem is dedicated to constructing a dynamic visual memory update mechanism, thereby maximizing the amount of information in streaming video within a limited memory space, much like the human brain. In practice, CausalMem estimates the redundancy of visual tokens and updates the memory bank via an online semantic basis, which models the principal semantics of the observed video stream. To validate CausalMem, we apply it to two representative MLLMs, namely LLaVA-OneVision and Qwen2.5-VL respectively, and conduct extensive experiments on both streaming and offline video understanding benchmarks. The experimental results not only show the great advantages than existing methods under both streaming and offline settings, \emph{e.g.}, $+3.2\%$ and $+3.0\%$ average accuracy gains respectively, but also witness the superior semantic preservation for streaming videos, \emph{e.g.}, using 12$k$ token budgets to memorize hour-long streaming videos, which achieves more than \textbf{20$\times$} visual token compression ratio and only occupies about \textbf{82 MB} storage. \textbf{Our code} is given in \href{https://github.com/hktk07/CausalMem}{CausalMem}.

cs.CV↗
arXiv2026

Ground Then Rank: Revisiting Knowledge-Based VQA with Training-Free Entity Identification

Knowledge-Based Visual Question Answering (KB-VQA) requires grounding visual queries to external knowledge beyond directly observable content in images. While recent multi modal large language models (MLLMs) show strong perceptual abilities, they struggle on KB-VQA tasks requiring groundings from both fine-grained entity and evidence levels. Most existing multi-modal retrieval augmented generation (MM-RAG) methods tightly couple entity discrimination and section-level evidence ranking into a single re-ranking stage, leading to high cost and limited generalization. In this work, we revisit existing MM-RAG solutions from a workflow perspective and argue both entity-level and fact-level groundings are key bottlenecks. We observe that although MLLMs often fail under open-ended entity naming, they can better identify the correct entity when selecting from a small set of candidate names. Based on this insight, we propose a simple and training-free identify-before-answer IBA framework that decouples entity identification from section-level re-ranking. Our approach prompts an MLLM to select high-confidence entities using only candidate names, followed by an off-the-shelf textual re-ranker for evidence selection. Experiments on Encyclopedic-VQA and InfoSeek show that our method consistently outperforms fine-tuned multi-modal re-ranking baselines while reducing training and inference complexity. Additional analyses reveal that the improvements arise not only from better entity identification, but also from selecting more informative evidence once correct entity is fixed. Our implementation is made public to ease reproducibility.

cs.CL↗
arXiv2026

UME: A Unified Meta-Generalization Framework for Cross-Domain ETA

Accurate Estimated Time of Arrival (ETA) prediction on checkout page is crucial in instant logistics for enhancing user satisfaction, optimizing dispatching, and controlling operational costs. In international on-demand delivery platforms, where ETA data originates from diverse countries or regions with different patterns, multi-domain modeling is of great importance and has been widely adopted. However, existing methods still face three critical challenges in real-world deployment. First, current multi-domain models struggle to generalize to completely unseen domains, failing to achieve zero-shot prediction during the initial cold-start phase. Second, cross-domain feature spaces are often assumed to be consistent, whereas new domains commonly suffer from structural missingness of offline (statistical) features due to the lack of historical data. Third, such feature missingness often compels industrial systems to model mature and cold-start domains separately, hindering knowledge transfer and increasing maintenance overhead. To address these challenges, we propose \textbf{UME}, a \textbf{U}nified \textbf{M}eta-generalization framework for \textbf{E}TA. Specifically, UME integrates a unified dual-branch architecture with a novel meta-learning mechanism that employs a hypernetwork-based meta learner. By leveraging domain-level knowledge and instance-level context, the meta learner empowers three meta modules to dynamically modulate feature gating, expert attention, and final prediction, capturing cross-domain correlations and facilitating intra-domain adaptation. A knowledge distillation strategy is further introduce to enhance performance. UME has now been deployed in Meituan-keeta delivery platform (the largest international food delivery platform in China). Extensive offline experiments and online A/B tests demonstrate that UME significantly outperforms existing baselines.

cs.LG↗
arXiv2026

Unifying and Optimizing Data Values for Selection via Sequential Decision-Making

Data selection has emerged as a crucial downstream application of data valuation, yet the theoretical foundations for using data values in selection remain underexplored. We reformulate data selection as a sequential decision-making problem where the optimal selection sequence arises from dynamic programming, and data values can be understood as encodings of this optimal sequence. This framework unifies and reinterprets existing methods like Data Shapley through the lens of approximate dynamic programming, revealing them as myopic linear approximations to the sequential problem. We further analyze how selection optimality degrades with utility curvature under submodularity, explaining when and why these approximations fail. To bridge theory and practice, we propose an efficient bipartite graph-based surrogate that preserves submodular structure while enabling scalable greedy selection with provable guarantees. Experiments on classical ML benchmarks and large-scale LLM fine-tuning data selection demonstrate substantial improvements over existing methods. Code is publicly available at https://github.com/frankhlchi/SeqDataVal

cs.AI↗
arXiv2026

SOMA: Efficient Multi-turn LLM Serving via Small Language Model

Large Language Models (LLMs) are increasingly deployed in multi-turn dialogue settings where preserving conversational context across turns is essential. A standard serving practice concatenates the full dialogue history at every turn, which reliably maintains coherence but incurs substantial cost in latency, memory, and API expenditure, especially when queries are routed to large proprietary models. Existing approaches often struggle to balance the trade-off between response quality and efficiency. We propose a framework that exploits the early turns of a session to estimate a local response manifold and then adapt a smaller surrogate model to this local region for the remainder of the conversation. Concretely, we learn soft prompts that maximize semantic divergence between the large and surrogate small language models' responses to surface least-aligned local directions, stabilize training with anti-degeneration control, and distill the mined cases into localized LoRA fine-tuning so the surrogate runs without prompts at inference. A simple gate enables a one-time switch with rollback on drift. We further provide a theoretical analysis for key components in SOMA. Extensive experiments show the effectiveness of SOMA. The source code is provided at: https://github.com/LabRAI/SOMA.

cs.CL↗
arXiv2026

Joint Optimization of Trajectory Control, Resource Allocation, and Task Offloading for Multi-UAV-Assisted IoV

This paper investigates a multi-Unmanned Aerial Vehicle (UAV) joint base station-assisted Internet of Vehicles (IoV) task offloading system in dense urban environments. To minimize system delay and energy consumption under strict coupling constraints, the complex non-convex optimization problem is decoupled into a hierarchical execution framework. First, a sequential distributed optimization algorithm based on Second-Order Cone Programming (SOCP) is proposed to optimize the 3D flight trajectory of each UAV, ensuring adaptive network coverage. Second, a novel hybrid resource scheduling paradigm synergizing Deep Reinforcement Learning (DRL) and Large Language Models (LLMs) is developed. Within this framework, the DRL agent dictates the initial resource allocation, while the LLM acts as a semantic macro-scheduler to rectify long-tail allocation imbalances for failed and surplus tasks. Crucially, a reward decoupling mechanism is introduced to isolate DRL training from external LLM interventions, thereby ensuring policy convergence. Finally, the task offloading ratios are precisely determined via Linear Programming (LP) within an alternating optimization loop. Simulation results demonstrate that the proposed method significantly outperforms traditional multi-agent reinforcement learning baselines in terms of task success rate and system efficiency.

cs.NI↗
arXiv2026

Single-Step Six-Dimensional Movable Antenna Reconfiguration for High-Mobility IoV: Modeling, Analysis, and Optimization

The Six-Dimensional Movable Antenna (6DMA) system has emerged as a promising technology to enhance wireless capacity by fully exploiting spatial degrees of freedom. However, applying 6DMA to high-mobility Internet of Vehicles (IoV) scenarios faces significant challenges, primarily due to the difficulty of acquiring instantaneous Channel State Information (CSI) and the risk of service interruptions caused by mechanical reconfiguration delays. To address these issues, this paper proposes a low-complexity, CSI-free single-step reconfiguration framework. First, we design a deterministic discrete position generation scheme based on a latitude-longitude grid with inherent topological structures. Leveraging graph theory, we explicitly model and theoretically derive the lower bounds of movement and time costs for antenna reconfiguration. Subsequently, utilizing the directional sparsity of 6DMA channels, we develop an adaptive optimization strategy that fuses offline environmental priors with online historical feedback. Furthermore, a periodic reconfiguration mechanism based on predicted cumulative vehicle distributions is introduced. By strictly restricting antenna adjustments to the first-order spatial neighborhood, the proposed single-step method effectively eliminates service interruptions. Simulation results demonstrate that the proposed scheme significantly outperforms traditional fixed and global-search-based benchmarks in terms of uplink sum rate, while incurring negligible mechanical overhead and latency, thereby validating its feasibility and robustness in highly dynamic vehicular networks.

cs.NI↗
arXiv2026

SpecSyn: LLM-based Synthesis and Refinement of Formal Specifications for Real-world Program Verification

Program verification is a formal technique to rigorously ensure the correctness and fault-freeness of software systems. However, constructing comprehensive interprocedural specifications for full verification obligations is time-consuming and labor-intensive, giving rise to automated specification generation approaches. Despite the significant advancements in these approaches brought by Large Language Models (LLMs), existing LLM-empowered approaches still suffer from significant limitations: they lack effective strategies for handling sizable input programs, and are typically equipped with no mechanisms to evaluate and guarantee the strength of the generated specifications. The limitations impair their ability to extract precise specifications from real-world complicated programs to support the verification of target properties, thereby hindering the applicability of existing approaches in verification tasks on real-world programs. To remedy this gap, we propose SpecSyn, a novel LLM-based specification generation method. SpecSyn first decomposes the input program into individual segments, which are handled respectively by the subsequent iterative specification generation process. Innovatively, we incorporate into the process a specification refinement mechanism based on semantic-non-equivalent program mutations and variant discrimination, assessing and enhancing the semantic strength of the generated specifications. Extensive experiments show that SpecSyn maintains high precision over 90% and outstanding recall over 75%, significantly outperforming existing LLM-based approaches. In further evaluations, SpecSyn successfully handles 1071 out of 1365 target properties for open-source programs, proving its applicability on real-world program verification tasks.

cs.SE↗
arXiv2026

ForestPrune: High-ratio Visual Token Compression for Video Multimodal Large Language Models via Spatial-Temporal Forest Modeling

Due to the great saving of computation and memory overhead, token compression has become a research hot-spot for MLLMs and achieved remarkable progress in image-language tasks. However, for the video, existing methods still fall short of high-ratio token compression. We attribute this shortcoming to the insufficient modeling of temporal and continual video content, and propose a novel and training-free token pruning method for video MLLMs, termed ForestPrune, which achieves effective and high-ratio pruning via Spatial-temporal Forest Modeling. In practice, ForestPrune construct token forests across video frames based on the semantic, spatial and temporal constraints, making an overall comprehension of videos. Afterwards, ForestPrune evaluates the importance of token trees and nodes based on tree depth and node roles, thereby obtaining a globally optimal pruning decision. To validate ForestPrune, we apply it to two representative video MLLMs, namely LLaVA-Video and LLaVA-OneVision, and conduct extensive experiments on a bunch of video benchmarks. The experimental results not only show the great effectiveness for video MLLMs, e.g., retaining 95.8% average accuracy while reducing 90% tokens for LLaVA-OneVision, but also show its superior performance and efficiency than the compared token compression methods, e.g., +10.1% accuracy on MLVU and -81.4% pruning time than FrameFusion on LLaVA-Video.

cs.CV↗
arXiv2026

Towards Effective Long Video Understanding of Multimodal Large Language Models via One-shot Clip Retrieval

Due to excessive memory overhead, most Multimodal Large Language Models (MLLMs) can only process videos of limited frames. In this paper, we propose an effective and efficient paradigm to remedy this shortcoming, termed One-shot video-Clip based Retrieval-Augmented Generation (OneClip-RAG). Compared with existing video RAG methods, OneClip-RAG makes full use of the merits of video clips for augmented video understanding in terms of both knowledge integrity and semantic coherence. Besides, it is also equipped with a novel query-guided video chunking algorithm that can unify clip chunking and cross-modal retrieval in one processing step, avoiding redundant computations. To improve instruction following, we further propose a new dataset called SynLongVideo and design a progressive training regime for OneClip-RAG. OneClip-RAG is plugged into three recent MLLMs and validated on a set of long-video benchmarks. Experimental results not only show the obvious performance gains by OneClip-RAG over MLLMs, e.g., boosting Qwen3-VL 8B to the level of GPT-5 on MLVU, but also show its superior efficiency in handling long videos. e.g., enabling LLaVA-Video understand up to an hour of videos in less than 1.2 minutes on a single 4090 GPU.

cs.CV↗