Cloud-based large language models (LLMs) have gained widespread adoption among human users. However, when the users shift to Internet-enabled machines, centralized LLM systems often suffer from high latency and fail to provide timely responses. Deploying LLMs over edge networks presents a promising alternative, yet maintaining an up-to-date knowledge base to address the dynamic and time-sensitive demands of diverse machine users remains a significant challenge. Consequently, there is a pressing need for fast and cost-efficient LLM deployment strategies tailored to edge environments. In this work, we address the problem of deploying mixture-of-experts (MoE) LLMs in a runtime- and cost-efficient manner. We formally define the Expert Model Deployment in Edge Networks (EMD-EN) problem, aiming to minimize deployment costs. Leveraging the inherent modularity of MoE, we propose a novel Neighbor-First Centrality (NFC) metric to facilitate the placement of model components across edge nodes and design the NFC-based Mixture of Expert layer Deployment (NFC-MoED) algorithm. Our results show that NFC-MoED substantially improves runtime efficiency and maintains near-optimal deployment costs compared to the brute-force benchmark.

The Internet of Vehicles (IoV) is an emerging technology that enhances traffic security and transportation efficiency by enabling smart, connected vehicles to communicate and exchange messages. IoV networks are a key component of intelligent transportation systems in smart cities. However, these networks are vulnerable to malicious vehicles that disseminate deceptive messages or impersonate legitimate entities, which compromises network security. These adversarial vehicles jeopardize the integrity and availability of the IoV network, exposing it to various security threats, including both insider and outsider attacks. Such attacks can severely undermine the trust and reliability of communication between legitimate vehicles. To address these challenges, we propose TMSU-IoV, a UAV-assisted trust management scheme that integrates identity authentication technique and trust evaluation mechanism to ensure secure communication of connected vehicles in IoV networks. To counteract outsider attacks, we introduce a certificateless signature-based authentication method that guarantees the authenticity of messages exchanged between vehicles and UAVs. To mitigate insider threats, we propose a quality of service (QoS)-based trust evaluation mechanism. This mechanism consists of a prior trust evaluation method and a posterior trust evaluation method, designed to enhance both the credibility and timeliness of trust evaluation for connected vehicles. Formal security analysis confirms that the TMSU-IoV effectively resists a variety of insider and outsider attacks. Performance evaluation experiments demonstrate that the TMSU-IoV can accurately assess the trust levels of connected vehicles and outperform traditional trust evaluation methods.

Website Fingerprinting (WF) aims to identify users’ visited websites from encrypted traffic traces, disabling the anonymity of encrypted communication like the Tor network. It is practical to use historically labeled (source) data, e.g., public datasets, to pre-train a WF model, and then collect few incoming (target) data to re-train this model within a low cost. Unfortunately, there is always a considerable difference of latent feature distributions between the source and target data (i.e., the cross-domain problem) and an inevitable bias of feature distribution caused by a limited volume of target data (i.e., the biased distribution problem). Although current Few-Shot Learning-based WF (FSWF) methods achieve satisfactory performance on the efficient establishment, they lack cross-domain transferability, and meanwhile, are unable to alleviate the distribution bias. In this paper, we first systematically analyze the cross-domain problem among different domains of traffics, revealing the ubiquity and dominant factors of it. To mitigate the cross-domain and biased distribution problems, we propose a Distribution Calibrated Website Fingerprinting (DCWF) method that incorporates a two-stage distribution calibration process and a tailored circle network. In the two-stage calibration process, we first devise a re-modeling mechanism capturing the information distribution of the target domain to extract representative features, and then design a calibration process to adjust the biased distribution of the target domain. Subsequently, a tailored circle network is proposed to reduce the noise caused by the calibration process. Finally, extensive experiments are conducted and the results demonstrate the superiority of our DCWF over comparisons under both close-world and open-world settings.

With rapidly increasing drug abuse across the world, it is imperative to monitor their supply chain with sufficient transparency. Blockchain is a common solution for achieving transparency in supply chain monitoring, but it does not have sufficient throughput for large-scale supply chains. It is challenging to achieve throughput and privacy simultaneously because complex dependencies among the supply chain events and the need for aggregation both make the application of ZKP challenging. We present PrivHChain, a privacy-preserving hierarchical blockchain that preserves transaction privacy even against blockchain peers while allowing them to verify record consistencies. This is enabled by novel modeling of supply chain events which makes it possible to use novel efficient zero-knowledge protocol schemes to verify the complex dependencies. Novel aggregation techniques are proposed to enable the proof aggregation, and the proofs are used to design monitoring protocols. PrivHChain is implemented and validated with extensive experiments and simulations. The results indicate that (i) the extra overhead of encryption and ZKP schemes is acceptable or negligible, and (ii) the throughput is improved by up to 5 times in simulations even with all the encryption/ZKP schemes.

The pervasiveness of offensive language on social networks has caused adverse effects on society, such as abusive behavior online. It is urgent to detect offensive language and curb its spread. In the popular datasets, the distribution of users and tweets is imbalanced, which limits the generalization ability of the model. In addition, existing research shows that methods with community information extracted from the social graphs effectively improve the performance of offensive language detection. However, the existing models deal with social graphs independently, which seriously affects the effectiveness of detection models. In this article, we release a new dataset with users and social relationships. To encode community information, we construct the social graphs based on the user historical behavior information and social relationships. Moreover, we propose a model based on graph attention networks (GATs) and fusion features for offensive language detection (GF-OLD). Specifically, the community information is directly captured by the GAT module, and the text embeddings are taken from the last hidden layer of bidirectional encoder representation from transformer (BERT). Attention mechanisms and position encoding are used to fuse these features. Our method outperforms baselines with the F1-score of 89.94%. The results show that our model effectively learns the potential information of social graphs and text, and user historical behavior information is more suitable for user attribute in the social graphs.

Online data trading is increasingly prevalent as data are becoming valuable assets. In most common conventional data trading scenarios, three parties (seller, broker, and buyer) exist, and fairness in trading is essential. This paper discusses and solves the fairness problem in two aspects. First, we considerexchange fairness, which requires payments and data exchanged correctly between buyers and the broker. In existing solutions, keys of encrypted data are traded. However, these solutions failed to provide a complete and secure design for validating keys' correctness unless they used generic theoretical but expensive methods, e.g., zk-SNARK. We address this security issue by designing a new key verification mechanism. We also present a novel atomic exchange protocol based on Hashed Timelock Contracts on Ethereum, reducing gas consumption compared to the existing approach. Second, we considerdistribution fairness, which requires correctly splitting income between the broker and sellers. Straightforward solutions are impractical, i.e., sellers participating in every transaction or traversing the blockchain. Therefore, we design a verifiable statement protocol for sellers to verify the income split efficiently. Further, analysis and experimental results indicate that extra fairness properties are securely achieved, and our protocol reduces users' on-chain participation compared to state-of-the-art protocols.

Frequent cyber-attacks force the service provider to employ security-aware service functions (SFs) to accommodate client network requests. Thanks to virtualization techniques' maturity, a security-aware SF can be provided by diverse vendors with various configurations, each of which needs various implementation cost and provides different security levels. When a client's network request comes, the multi-configuration SFs could compose various security-aware service function chains (S-SFCs) to flexibly satisfy the security requirement. In this paper, we investigate how to efficiently compose and embed an S-SFC to satisfy the client's security requirement. With the objective of cost optimization, we formulate the problem of security-aware service function chain deployment and prove its NP-hardness. We propose the technique of the security-cost-balance (SCB) factor to efficiently consider the capability of a physical node and the cost when the node is employed to satisfy the client's security requirement. Based on this technique, we develop an efficient algorithm called SCB-based S-SFC deployment (SCB-SD). The simulation results show that SCB-SD significantly outperforms the benchmarks directly extended from the state-of-the-art.

This paper presents an approach to enhance the quality of machine translation by leveraging middle sentences as pivot points and employing dual reinforcement learning. Conventional methods for generating parallel sentence pairs for machine translation rely on parallel corpora, which may be scarce, resulting in limitations in translation quality. In contrast, our proposed method entails training two machine translation models in opposite directions, utilizing the middle sentence as a bridge for a virtuous feedback loop between the two models. This feedback loop resembles reinforcement learning, facilitating the models to make informed decisions based on mutual feedback. Experimental results substantiate that our proposed method significantly improves machine translation quality.

Local differential privacy has been considered the standard measurement for privacy preservation in distributed data collection. Corresponding mechanisms have been designed for multiple types of tasks, like the frequency estimation for categorical values and the mean value estimation for numerical values. However, the histogram publication of numerical values, containing abundant and crucial clues for the whole dataset, has not been thoroughly considered under this measurement. To simply encode data into different intervals upon each query will soon exhaust the bandwidth and the privacy budgets, which is infeasible for real scenarios. Therefore, this paper proposes a highly efficient framework for differentially private histogram publication of numerical values in a distributed environment. The proposed algorithms can efficiently adopt the correlations among multiple queries and achieve an optimal resource consumption. We also conduct extensive experiments on real‐world data traces, and the results validate the improvement of proposed algorithms.

Nowadays, the social network becomes an indispensable part of people's daily life, meanwhile offers an unprecedentedly convenient access for purposive individuals to influence the opinions of network users. Current studies present a subtle persuasion approach that finds a number of key users meanwhile varies their susceptibility extent to impact the public opinion. Such persuasion is significantly critical for public security, as it could facilitate both the spreading and dispelling of malicious rumors. However, the major body of these studies enclose impractical assumptions, such that persuaders have an unlimited budget, or the costs of varying different users' susceptibilities are the same, thus rendering these works unsuitable for realistic scenarios. Therefore, this work originally proposes a more practical and generalized problem of persuasion, where varying the susceptibilities of different users holds different costs. The analysis of its non-convexity, non-submodularity and complexity shows that solving the proposed problem is nontrivial, thus inspiring us to provide an intuitive greedy algorithm. Furthermore, we design an accelerated algorithm based on the community property, which reduces the time consumption more than one order of magnitude. The acceleration is based on the intuition that the impact of a user within a proper community could be a good estimation of the impact in the whole network, while the computation of the former one is much more efficient. The relationship between two algorithms is fully analyzed, which shows the community-based algorithm can degenerate to the intuitive greedy algorithm under a specific setting. Finally, comprehensive evaluations on real-world datasets show the superiority of proposed algorithms on both effectiveness and efficiency.

In social networks, several influential individuals can promote an idea or a product to numerous individuals. Thus, it is valuable to solve the influence maximization (IM) problem, which asks for finding the most influential set of individuals in a social network. To estimate the influence of individuals, the existing independent cascade (IC) model simulates the influence diffusion only considering the influences from direct in-neighbors to nodes. This consideration does not hold in real life. In many cases, people are likely influenced by information depending on where it comes from, instead of who gives it. To simulate the influence diffusion more accurate, this paper proposes the second-order IC model, which takes the previous influence into consideration. In addition, we design an approximate algorithm and its distributed extension for IM under the second-order IC model. Experimental results show that our second-order IC model outperforms the IC model in terms of simulating influence diffusions. The proposed algorithms are efficient, and the obtained node sets are influential.

The Internet of Things (IoT) has attracted the interests of both academia and industry and enables various real-world applications. The acquirement of large amounts of sensing data is a fundamental issue in IoT. An efficient way is obtaining sufficient data by the mobile crowdsensing. It is a promising paradigm which leverages the sensing capacity of portable mobile devices. The crowdsensing platform is the key entity who allocates tasks to participants in a mobile crowdsensing system. The strategy of task allocating is crucial for the crowdsensing platform, since it affects the data requester’s confidence, the participant’s confidence, and its own benefit. Traditional allocating algorithms regard the privacy preservation, which may lose the confidence of participants. In this paper, we propose a novel three-step algorithm which allocates tasks to participants with privacy consideration. It maximizes the benefit of the crowdsensing platform and meanwhile preserves the privacy of participants. Evaluation results on both benefit and privacy aspects show the effectiveness of our proposed algorithm.

Recently, CNN (Convolutional Neural Network) based trackers have achieved promising results benefited from their robust feature representation. However, most trackers only use features from a certain layer, which limits their performance. In this paper, we propose a novel CNN based tracker. Firstly, we use local detection and global detection network for target localization. In local detection network, we fuse features from different layers to train a fully convolutional neural network for target localization. In case the local detection network fails when the target disappear for a while and appears in another location, we train a global detection network to detect if the target appears again. Then, we employ a correlation filter to estimate accurate scale of the target using HOG features extracted around predicted location. Extensive experiments on various challenging video sequences demonstrate the effectiveness of our proposed algorithm compared with several state-of-the-art trackers.