arXiv subjects

Yijing Lin

Publications and source records attributed to Yijing Lin.

16 recordsLinked to original sources

arXiv2026

CBD: API-Only LLM Black-Box Unlearning through Controlled Behavioral Divergence

Edge devices increasingly invoke large language models (LLMs) through API services for context aware edge intelligence, while edge generated data may be collected to improve LLMs and may introduce sensitive, copyrighted, harmful, or outdated information into model behavior. Machine unlearning offers a practical way to remove the influence of undesired data without retraining LLMs. However, existing methods still face two gaps. The first is API only black box access, where target model parameters and internal logits are unavailable. The second is how to preserve retained utility when unlearning target data and retained data share highly similar prompt structures or semantic patterns. To address these challenges, we propose Controlled Behavioral Divergence (CBD), an API only black box unlearning framework. CBD uses two auxiliary models to create controlled behavioral divergence between retained inputs and unlearning target inputs, converts this divergence into an unlearning relevance score, and routes unlearning related prompts away from the target LLM. To improve discrimination accuracy under high similarity between target and retained data, CBD constructs a gradient statistics based discriminative basis by estimating empirical Fisher matrices and solving a regularized generalized eigenvalue problem, guiding the unlearning signal toward target specific information rather than shared prompt structures. Compared with eleven white box and gray box unlearning baselines, CBD achieves a better unlearning utility trade off and its performance varies little across settings. On ToFU forget10, CBD approaches the retrained reference on the forget set while raising model utility to 74.90, about 15% above the second best baseline. On WMDP, it lowers hazardous knowledge accuracy to 25.68, near random guessing, while preserving MMLU accuracy of 52.67. Code is at https://github.com/DGL-codes/CBD.

cs.LG↗

arXiv2026

Intellicise Wireless Networks Meet Agentic AI: A Security and Privacy Perspective

Intellicise (Intelligent and Concise) wireless network is the main direction of the evolution of future mobile communication systems, a perspective now widely acknowledged across academia and industry. As a key technology within it, Agentic AI has garnered growing attention due to its advanced cognitive capabilities, enabled through continuous perception-memory-reasoning-action cycles. This paper first analyses the unique advantages that Agentic AI introduces to intellicise wireless networks. We then propose a structured taxonomy for Agentic AI-enhanced secure intellicise wireless networks. Building on this framework, we identify emerging security and privacy challenges introduced by Agentic AI and summarize targeted strategies to address these vulnerabilities. A case study further demonstrates Agentic AI's efficacy in defending against intelligent eavesdropping attacks. Finally, we outline key open research directions to guide future exploration in this field.

cs.CR↗

arXiv2025

Secure Multi-LLM Agentic AI and Agentification for Edge General Intelligence by Zero-Trust: A Survey

Agentification serves as a critical enabler of Edge General Intelligence (EGI), transforming massive edge devices into cognitive agents through integrating Large Language Models (LLMs) and perception, reasoning, and acting modules. These agents collaborate across heterogeneous edge infrastructures, forming multi-LLM agentic AI systems that leverage collective intelligence and specialized capabilities to tackle complex, multi-step tasks. However, the collaborative nature of multi-LLM systems introduces critical security vulnerabilities, including insecure inter-LLM communications, expanded attack surfaces, and cross-domain data leakage that traditional perimeter-based security cannot adequately address. To this end, this survey introduces zero-trust security of multi-LLM in EGI, a paradigmatic shift following the ``never trust, always verify'' principle. We begin by systematically analyzing the security risks in multi-LLM systems within EGI contexts. Subsequently, we present the vision of a zero-trust multi-LLM framework in EGI. We then survey key technical progress to facilitate zero-trust multi-LLM systems in EGI. Particularly, we categorize zero-trust security mechanisms into model- and system-level approaches. The former and latter include strong identification, context-aware access control, etc., and proactive maintenance, blockchain-based management, etc., respectively. Finally, we identify critical research directions. This survey serves as the first systematic treatment of zero-trust applied to multi-LLM systems, providing both theoretical foundations and practical strategies.

cs.NI↗

arXiv2025

Label Inference Attacks against Federated Unlearning

Federated Unlearning (FU) has emerged as a promising solution to respond to the right to be forgotten of clients, by allowing clients to erase their data from global models without compromising model performance. Unfortunately, researchers find that the parameter variations of models induced by FU expose clients' data information, enabling attackers to infer the label of unlearning data, while label inference attacks against FU remain unexplored. In this paper, we introduce and analyze a new privacy threat against FU and propose a novel label inference attack, ULIA, which can infer unlearning data labels across three FU levels. To address the unique challenges of inferring labels via the models variations, we design a gradient-label mapping mechanism in ULIA that establishes a relationship between gradient variations and unlearning labels, enabling inferring labels on accumulated model variations. We evaluate ULIA on both IID and non-IID settings. Experimental results show that in the IID setting, ULIA achieves a 100% Attack Success Rate (ASR) under both class-level and client-level unlearning. Even when only 1% of a user's local data is forgotten, ULIA still attains an ASR ranging from 93% to 62.3%.

cs.CR↗

arXiv2025

Large AI Model-Enabled Secure Communications in Low-Altitude Wireless Networks: Concepts, Perspectives and Case Study

Low-altitude wireless networks (LAWNs) have the potential to revolutionize communications by supporting a range of applications, including urban parcel delivery, aerial inspections and air taxis. However, compared with traditional wireless networks, LAWNs face unique security challenges due to low-altitude operations, frequent mobility and reliance on unlicensed spectrum, making it more vulnerable to some malicious attacks. In this paper, we investigate some large artificial intelligence model (LAM)-enabled solutions for secure communications in LAWNs. Specifically, we first explore the amplified security risks and important limitations of traditional AI methods in LAWNs. Then, we introduce the basic concepts of LAMs and delve into the role of LAMs in addressing these challenges. To demonstrate the practical benefits of LAMs for secure communications in LAWNs, we propose a novel LAM-based optimization framework that leverages large language models (LLMs) to generate enhanced state features on top of handcrafted representations, and to design intrinsic rewards accordingly, thereby improving reinforcement learning performance for secure communication tasks. Through a typical case study, simulation results validate the effectiveness of the proposed framework. Finally, we outline future directions for integrating LAMs into secure LAWN applications.

cs.NI↗

arXiv2025

Toward Low-Altitude Airspace Management and UAV Operations: Requirements, Architecture and Enabling Technologies

The low-altitude economy (LAE) is rapidly advancing toward intelligence, connectivity, and coordination, bringing new challenges in dynamic airspace management, unmanned aerial vehicle (UAV) operation, and security management. Existing systems remain fragmented and lack effective coordination. To bridge these gaps, we propose UTICN (Ubiquitous and Trusted Intelligent Cellular-native Network) for LAE, a unified cellular-native architecture that integrates multi-domain sensing, high-precision positioning, intelligent aircraft-to-everything communication, dynamic airspace management, and UAV operational services. UTICN introduces key technologies such as integrated sensing and communication (ISAC), passive and active positioning, intelligent machine communication, swarm coordination, and control-data decoupled management frameworks. We demonstrate UTICN's feasibility through two use cases, i.e., a city-level LAE management platform and a multi-frequency collaborative ISAC system. This work provides a fundamental reference for building a unified operational foundation and airspace management architecture for the LAE.

eess.SY↗

arXiv2025

RealGeneral: Unifying Visual Generation via Temporal In-Context Learning with Video Models

Unifying diverse image generation tasks within a single framework remains a fundamental challenge in visual generation. While large language models (LLMs) achieve unification through task-agnostic data and generation, existing visual generation models fail to meet these principles. Current approaches either rely on per-task datasets and large-scale training or adapt pre-trained image models with task-specific modifications, limiting their generalizability. In this work, we explore video models as a foundation for unified image generation, leveraging their inherent ability to model temporal correlations. We introduce RealGeneral, a novel framework that reformulates image generation as a conditional frame prediction task, analogous to in-context learning in LLMs. To bridge the gap between video models and condition-image pairs, we propose (1) a Unified Conditional Embedding module for multi-modal alignment and (2) a Unified Stream DiT Block with decoupled adaptive LayerNorm and attention mask to mitigate cross-modal interference. RealGeneral demonstrates effectiveness in multiple important visual generation tasks, e.g., it achieves a 14.5% improvement in subject similarity for customized generation and a 10% enhancement in image quality for canny-to-image task. Project page: https://lyne1.github.io/realgeneral_web/; GitHub Link: https://github.com/Lyne1/RealGeneral

cs.CV↗

arXiv2024

Mixture of Experts for Network Optimization: A Large Language Model-enabled Approach

Optimizing various wireless user tasks poses a significant challenge for networking systems because of the expanding range of user requirements. Despite advancements in Deep Reinforcement Learning (DRL), the need for customized optimization tasks for individual users complicates developing and applying numerous DRL models, leading to substantial computation resource and energy consumption and can lead to inconsistent outcomes. To address this issue, we propose a novel approach utilizing a Mixture of Experts (MoE) framework, augmented with Large Language Models (LLMs), to analyze user objectives and constraints effectively, select specialized DRL experts, and weigh each decision from the participating experts. Specifically, we develop a gate network to oversee the expert models, allowing a collective of experts to tackle a wide array of new tasks. Furthermore, we innovatively substitute the traditional gate network with an LLM, leveraging its advanced reasoning capabilities to manage expert model selection for joint decisions. Our proposed method reduces the need to train new DRL models for each unique optimization problem, decreasing energy consumption and AI model implementation costs. The LLM-enabled MoE approach is validated through a general maze navigation task and a specific network service provider utility maximization task, demonstrating its effectiveness and practical applicability in optimizing complex networking systems.

cs.NI↗

arXiv2024

Blockchain-enabled Trustworthy Federated Unlearning

Federated unlearning is a promising paradigm for protecting the data ownership of distributed clients. It allows central servers to remove historical data effects within the machine learning model as well as address the "right to be forgotten" issue in federated learning. However, existing works require central servers to retain the historical model parameters from distributed clients, such that allows the central server to utilize these parameters for further training even, after the clients exit the training process. To address this issue, this paper proposes a new blockchain-enabled trustworthy federated unlearning framework. We first design a proof of federated unlearning protocol, which utilizes the Chameleon hash function to verify data removal and eliminate the data contributions stored in other clients' models. Then, an adaptive contribution-based retraining mechanism is developed to reduce the computational overhead and significantly improve the training efficiency. Extensive experiments demonstrate that the proposed framework can achieve a better data removal effect than the state-of-the-art frameworks, marking a significant stride towards trustworthy federated unlearning.

cs.LG↗

arXiv2024

Scalable Federated Unlearning via Isolated and Coded Sharding

Federated unlearning has emerged as a promising paradigm to erase the client-level data effect without affecting the performance of collaborative learning models. However, the federated unlearning process often introduces extensive storage overhead and consumes substantial computational resources, thus hindering its implementation in practice. To address this issue, this paper proposes a scalable federated unlearning framework based on isolated sharding and coded computing. We first divide distributed clients into multiple isolated shards across stages to reduce the number of clients being affected. Then, to reduce the storage overhead of the central server, we develop a coded computing mechanism by compressing the model parameters across different shards. In addition, we provide the theoretical analysis of time efficiency and storage effectiveness for the isolated and coded sharding. Finally, extensive experiments on two typical learning tasks, i.e., classification and generation, demonstrate that our proposed framework can achieve better performance than three state-of-the-art frameworks in terms of accuracy, retraining time, storage overhead, and F1 scores for resisting membership inference attacks.

cs.LG↗

arXiv2023

Enhancing Deep Reinforcement Learning: A Tutorial on Generative Diffusion Models in Network Optimization

Generative Diffusion Models (GDMs) have emerged as a transformative force in the realm of Generative Artificial Intelligence (GenAI), demonstrating their versatility and efficacy across various applications. The ability to model complex data distributions and generate high-quality samples has made GDMs particularly effective in tasks such as image generation and reinforcement learning. Furthermore, their iterative nature, which involves a series of noise addition and denoising steps, is a powerful and unique approach to learning and generating data. This paper serves as a comprehensive tutorial on applying GDMs in network optimization tasks. We delve into the strengths of GDMs, emphasizing their wide applicability across various domains, such as vision, text, and audio generation. We detail how GDMs can be effectively harnessed to solve complex optimization problems inherent in networks. The paper first provides a basic background of GDMs and their applications in network optimization. This is followed by a series of case studies, showcasing the integration of GDMs with Deep Reinforcement Learning (DRL), incentive mechanism design, Semantic Communications (SemCom), Internet of Vehicles (IoV) networks, etc. These case studies underscore the practicality and efficacy of GDMs in real-world scenarios, offering insights into network design. We conclude with a discussion on potential future directions for GDM research and applications, providing major insights into how they can continue to shape the future of network optimization.

cs.NI↗

arXiv2023

A Unified Framework for Integrating Semantic Communication and AI-Generated Content in Metaverse

As the Metaverse continues to grow, the need for efficient communication and intelligent content generation becomes increasingly important. Semantic communication focuses on conveying meaning and understanding from user inputs, while AI-Generated Content utilizes artificial intelligence to create digital content and experiences. Integrated Semantic Communication and AI-Generated Content (ISGC) has attracted a lot of attentions recently, which transfers semantic information from user inputs, generates digital content, and renders graphics for Metaverse. In this paper, we introduce a unified framework that captures ISGC two primary benefits, including integration gain for optimized resource allocation and coordination gain for goal-oriented high-quality content generation to improve immersion from both communication and content perspectives. We also classify existing ISGC solutions, analyze the major components of ISGC, and present several use cases. We then construct a case study based on the diffusion model to identify an optimal resource allocation strategy for performing semantic extraction, content generation, and graphic rendering in the Metaverse. Finally, we discuss several open research issues, encouraging further exploring the potential of ISGC and its related applications in the Metaverse.

cs.HC↗

arXiv2023

Blockchain-aided Secure Semantic Communication for AI-Generated Content in Metaverse

The construction of virtual transportation networks requires massive data to be transmitted from edge devices to Virtual Service Providers (VSP) to facilitate circulations between the physical and virtual domains in Metaverse. Leveraging semantic communication for reducing information redundancy, VSPs can receive semantic data from edge devices to provide varied services through advanced techniques, e.g., AI-Generated Content (AIGC), for users to explore digital worlds. But the use of semantic communication raises a security issue because attackers could send malicious semantic data with similar semantic information but different desired content to break Metaverse services and cause wrong output of AIGC. Therefore, in this paper, we first propose a blockchain-aided semantic communication framework for AIGC services in virtual transportation networks to facilitate interactions of the physical and virtual domains among VSPs and edge devices. We illustrate a training-based targeted semantic attack scheme to generate adversarial semantic data by various loss functions. We also design a semantic defense scheme that uses the blockchain and zero-knowledge proofs to tell the difference between the semantic similarities of adversarial and authentic semantic data and to check the authenticity of semantic data transformations. Simulation results show that the proposed defense method can reduce the semantic similarity of the adversarial semantic data and the authentic ones by up to 30% compared with the attack scheme.

cs.CR↗

arXiv2022

A Blockchain-based Semantic Exchange Framework for Web 3.0 toward Participatory Economy

Web 3.0 is the next-generation Internet that enables participants to read, write, and own contents in a decentralized manner. It is mainly driven by blockchain, semantic communication, edge computing, and artificial intelligence, which can construct value networks to achieve participatory economics based on participatory decision making. Web 3.0 can capture the characteristics of blockchain, semantic extraction, and communication to achieve decentralized semantic sharing and transfer information precisely. However, current Web 3.0 solutions focus on the blockchain while overlooking other new technologies' roles in Web 3.0. To further unleash the advantages of semantic extraction and communication in Web 3.0, in this paper, we propose a blockchain-based semantic exchange framework to realize fair and efficient interactions. In this framework, we first attempt to tokenize semantic information into Non-Fungible Token (NFT) for semantic exchange. Then we utilize a Stackelberg game to maximize buying and pricing strategies for semantic trading. We also leverage Zero-Knowledge Proof to share authentic semantic information without publishing it before receiving payments, which can achieve a fair and privacy-preserving trading compared with current NFT marketplaces. A case study about urban planning is given to show clearly the proposed mechanisms. Finally, several challenges and opportunities are identified.

cs.NI↗

arXiv2022

A Unified Blockchain-Semantic Framework for Wireless Edge Intelligence Enabled Web 3.0

Web 3.0 enables user-generated contents and user-selected authorities. With decentralized wireless edge computing architectures, Web 3.0 allows users to read, write, and own contents. A core technology that enables Web 3.0 goals is blockchain, which provides security services by recording content in a decentralized and transparent manner. However, the explosion of on-chain recorded contents and the fast-growing number of users cause increasingly unaffordable computing and storage resource consumption. A promising paradigm is to analyze the semantic information of contents that can convey precisely the desired meanings without consuming many resources. In this article, we propose a unified blockchain-semantic ecosystems framework for wireless edge intelligence-enabled Web 3.0. Our framework consists of six key components to exchange semantic demands. We then introduce an Oracle-based proof of semantic mechanism to implement on-chain and off-chain interactions of Web 3.0 ecosystems on semantic verification algorithms while maintaining service security. An adaptive Deep Reinforcement Learning-based sharding mechanism on Oracle is designed to improve interaction efficiency, which can facilitate Web 3.0 ecosystems to deal with varied semantic demands. Finally, a case study is presented to show that the proposed framework can dynamically adjust Oracle settings according to varied semantic demands.

cs.CR↗

arXiv2022

BcMON: Blockchain Middleware for Offline Networks

Blockchain is becoming a new generation of information infrastructures. However, the current blockchain solutions rely on a continuous connectivity network to query and modify the state of the blockchain. The emerging satellite technology seems to be a good catalyst to forward offline transactions to the blockchain. However, this approach suffers expensive costs, difficult interoperability, and limited computation problems. Therefore, we propose BcMON, the first blockchain middleware for offline networks. BcMON incorporates three innovative designs: 1) it reduces the costs of offline transactions accessing the blockchain through Short Message Service (SMS), 2) it validates the authenticity of offline cross-chain transactions by two-phase consensus, 3) it supports offline clients to perform complex queries and computations on the blockchains. The prototype of BcMON has been implemented to evaluate the performance of the proposed middleware, which can show its stability, efficiency, and scalability.

cs.CR↗