Deep Neural Networks (DNNs), as valuable intellectual property, face unauthorized use. Existing protections, such as digital watermarking, are largely passive; they provide only post-hoc ownership verification and cannot actively prevent the illicit use of a stolen model. This work proposes a proactive protection scheme, dubbed ``Authority Backdoor," which embeds access constraints directly into the model. In particular, the scheme utilizes a backdoor learning framework to intrinsically lock a model's utility, such that it performs normally only in the presence of a specific trigger (e.g., a hardware fingerprint). But in its absence, the DNN's performance degrades to be useless. To further enhance the security of the proposed authority scheme, the certifiable robustness is integrated to prevent an adaptive attacker from removing the implanted backdoor. The resulting framework establishes a secure authority mechanism for DNNs, combining access control with certifiable robustness against adversarial attacks. Extensive experiments on diverse architectures and datasets validate the effectiveness and certifiable robustness of the proposed framework.

Microcontroller units (MCUs) are widely used in embedded devices due to their low power consumption and cost-effectiveness. MCU firmware controls these devices and is vital to the security of embedded systems. However, performing dynamic security analyses for MCU firmware has remained challenging due to the lack of usable execution environments -- existing dynamic analyses cannot run on physical devices (e.g., insufficient computational resources), while building emulators is costly due to the massive amount of heterogeneous hardware, especially peripherals. Recent advances in automated peripheral emulation have made MCU emulation more scalable. However, these efforts only support limited peripherals and are hard to extend because they require ad-hoc adaptations.

The rapid evolution of the 6G network infrastructure has position cloud-edge systems integrating giant AI models and IoT devices as key enablers for next-generation surveillance solutions. However, traditional centralized architectures for such solutions face challenges in efficient multimodal data transmission and processing due to the massive data generated by terminal devices and the high computational demands of giant AI models served at the cloud side. To address these limitations, we propose DEVOLUTION, a 6G-based symbiotic cloud-edge framework that optimizes multimodal data transmission and the operation of the giant AI model in surveillance solutions. DEVOLUTION employs a two-stage hierarchical index encoding mechanism to dynamically distribute computation loads between Elasticsearch cluster edge devices and cloud servers, mitigating bandwidth constraints while preserving data locality. Furthermore, DEVOLUTION introduce a distributed giant model adaptation strategy, where cloud servers fine-tune the Chinese CLIP model (CN-CLIP), while edge devices deploy a lightweight Chinese CLIP Self-attention and Crossattention (CN-CLIP-SA-CA) fusion model, enabling secure and efficient cross-modal data transformation and alignment. In the end, a symbiotic retrieval engine, optimized with a best-first beam search (BFBS) strategy for 6G environments, ensures highaccuracy, low-latency multimodal retrieval. Experimental studies are being conducted on real datasets and the results demonstrate that DEVOLUTION significantly reduces latency, enhances privacy through edge-localized processing, and outperforms traditional methods (BM25, IVF, HIVF, KNN). A very small number of top candidate sets require encryption and decryption and transmission to recall. The retrieval precision can reach 96%–100% of the accuracy of Elasticsearch KNN, while retrieval reduces the retrieval time by 1.2%-45.2% compared to several latest schemes in different datasets and scales.

Website fingerprinting (WF) attack is a type of traffic analysis attack. It enables a local and passive eavesdropper situated between the Tor client and the Tor entry node to deduce which websites the client is visiting. Currently, deep learning (DL) based WF attacks have overcome a number of proposed WF defenses, demonstrating superior performance compared to traditional machine learning (ML) based WF attacks. To mitigate this threat, we present WFGuard, a fuzzing-testing-based traffic morphing WF defense technique. WFGuard employs fine-grained neuron information within WF classifiers to design a joint optimization function and then applies gradient ascent to maximize both neurons value and misclassification possibility in DL-based WF classifiers. During each traffic mutation cycle, we propose a gradient based dummy traffic injection pattern generation approach, continuously mutating the traffic until a pattern emerges that can successfully deceive the classifier. Finally, the pattern present in successful variant traces are extracted and applied as defense strategies to Tor traffic. Extensive evaluations reveal that WFGuard can effectively decrease the accuracy of DL-based WF classifiers (e.g., DF and Var-CNN) to a mere 4.43%, while only incurring an 11.04% bandwidth overhead. This highlights the potential efficacy of our approach in mitigating WF attacks.

For a web search engine, it is critical to design a mechanism to promote trustworthy websites and eliminate spam ones in the searching results. In this paper, we propose the OpinionRank algorithm to compute the trustworthiness of a website and identify trustworthy ones with high trust values. OpinionRank is essentially a breadth-first-search based algorithm that starts from an existing set of trustworthy websites, also called seeds. Because seeds play a vital role in OpinionRank, we put forward a novel seed selection scheme, named HarMean PageRank algorithm. HarMean combines the results of two seed selection algorithms, i.e. High PageRank and Inverse PageRank, to rank websites based on their trustworthiness. After trustworthy seeds are chosen, OpinionRank iteratively computes the trustworthiness of every website, leveraging trust propagation and trust combination. Using the public dataset WEBSPAM-UK2006, we validate OpinionRank and HarMean PageRank, analyze the impact of seed selection, and evaluate the convergence speed of OpinionRank. Experimental results indicate that OpinionRank can detect more trustworthy websites with fewer seeds, when compared to three state-of-the-art solutions, TrustRank, GoodRank, and Enhanced OpinionWalk algorithms.

Assessing trust in online social networks (OSNs) is critical for many applications such as online marketing and network security. It is a challenging problem, however, due to the difficulties of handling complex social network topologies and conducting accurate assessment in these topologies. To address these challenges, we model trust by proposing the three-valued subjective logic (3VSL) model. 3VSL properly models the uncertainties that exist in trust, thus is able to compute trust in arbitrary graphs. We theoretically prove the capability of 3VSL based on the Dirichlet-Categorical (DC) distribution and its correctness in arbitrary OSN topologies. Based on the 3VSL model, we further design the AssessTrust (AT) algorithm to accurately compute the trust between any two users connected in an OSN. We validate 3VSL against two real-world OSN datasets: Advogato and Pretty Good Privacy (PGP). Experimental results indicate that 3VSL can accurately model the trust between any pair of indirectly connected users in the Advogato and PGP.

Assessing the trust between users in a trust social network (TSN) isa critical issue in many applications, e.g., film recommendation,spam detection, and online lending. Despite of various trust assessment methods, a challenge remaining to existing solutions is how to accurately determine the factors that affect trust propagation and trust fusion within a TSN. To address this challenge, we propose the NeuralWalk algorithm to cope with trust factor estimation and trust relation prediction problems simultaneously. NeuralWalk employs a neural network, named WalkNet, to model single-hop trust propagation and fusion in a TSN. By treating original trust relations in a TSN as labeled samples, WalkNet is able to learn the parameters that will be used for trust computation/assessment. Unlike traditional solutions, WalkNet is able to accurately predict unknown trust relations in an inductive manner. Based on WalkNet, NeuralWalk iteratively assesses the unknown multi-hop trust relations among users via the obtained single-hop trust computation rules. Experiments on two real-world TSN datasets indicate that NeuralWalk significantly outperforms the state-of-the-art solutions.

Trustworthiness in a vehicular network plays a vital role in facilitating data sharing among vehicles to achieve better driving safety and convenience. Without trustworthiness assessment, a vehicle may not be able to trust other vehicles and, therefore, simply drop the data shared from others to avoid potential driving dangers. This problem was traditionally approached by protecting data security; however, the study of the trustworthiness of data generators (vehicles) is unfortunately omitted. We envision the existences of a vehicular social network on road, wherein vehicles exchanging data between each other are considered socially connected. Leveraging the trust propagation and fusion within a vehicular social network, the trustworthiness of individual vehicles can be accurately assessed. We adopt the three-valued subjective logic model to study trust between vehicles, and propose a holistic solution to trust assessment in vehicular social networks. The proposed solution enables objective and subjective trust assessment of vehicles, in a distributed manner. Simulation results indicate that the proposed solution offers a more accurate trust assessment and a quicker assessing time.

Water quality data is incredibly important and valuable, but its acquisition is not always trivial. A promising solution is to distribute a wireless sensor network in water to measure and collect the data; however, a drawback exists in that the batteries of the system must be replaced or recharged after being exhausted. To mitigate this issue, we designed a self-sustained water quality sensing system that is powered by renewable bioenergy generated from microbial fuel cells (MFCs). MFCs collect the energy released from native magnesium oxidizing microorganisms (MOMs) that are abundant in natural waters. The proposed energy-harvesting technology is environmentally friendly and can provide maintenance-free power to sensors for several years. Despite these benefits, an MFC can only provide microwatt-level power that is not sufficient to continuously power a sensor. To address this issue, we designed a power management module to accumulate energy when the input voltage is as low as 0.33V. We also proposed a radio-frequency (RF) activation technique to remotely activate sensors that otherwise are switched off in default. With this innovative technique, a sensor’s energy consumption in sleep mode can be completely avoided. Additionally, this design can enable on-demand data acquisitions from sensors. We implement the proposed system and evaluate its performance in a stream. In 3-month field experiments, we find the system is able to reliably collect water quality data and is robust to environment changes.

Artificial intelligence, and particularly machine learning, has been used in many ways by the research community to turn a variety of diverse and even heterogeneous data sources into high quality facts and knowledge, providing premier capabilities to accurate pattern discovery. However, applying machine learning strategies on big and complex datasets is computationally expensive, and it consumes a very large amount of logical and physical resources, such as data file space, CPU, and memory. A sophisticated platform for efficient big data analytics is becoming more important these days as the data amount generated in a daily basis exceeds over quintillion bytes. Apache Spark MLlib is one of the most prominent platforms for big data analysis which offers a set of excellent functionalities for different machine learning tasks ranging from regression, classification, and dimension reduction to clustering and rule extraction. In this contribution, we explore, from the computational perspective, the expanding body of the Apache Spark MLlib 2.0 as an open-source, distributed, scalable, and platform independent machine learning library. Specifically, we perform several real world machine learning experiments to examine the qualitative and quantitative attributes of the platform. Furthermore, we highlight current trends in big data machine learning research and provide insights for future work.

Massive trust assessment (MTA) in an Online Social Network (OSN), i.e., computing the trustworthiness of all users in the network, is crucial in various OSN-related applications. Existing solutions are either too slow or inaccurate in addressing the MTA problem. We propose the OpinionWalk algorithm that accurately and efficiently conducts MTA in an OSN. OpinionWalk models trust by the Dirichlet distribution and uses a matrix to represent the direct trust relations among users. From the perspective of a user, other users' trustworthiness are stored in a column vector that is iteratively updated when the algorithm “walks” through the network, in a breadth-first search manner. We identify the overlapping subproblems property in MTA and prove OpinionWalk is a more efficient solution. The accuracy and execution time of OpinionWalk are evaluated and compared to benchmark algorithms including EigenTrust, TrustRank, MoleTrust, TidalTrust and AssessTrust, using two real-world datasets (Advogato and Pretty Good Privacy). Experimental results indicate that OpinionWalk is an efficient and accurate solution to MTA, compared to previous algorithms.

Assessing multi-hop interpersonal trust in online social networks (OSNs) is critical for many social network applications such as online marketing but challenging due to the difficulties of handling complex OSN topology, in existing models such as subjective logic, and the lack of effective validation methods. To address these challenges, we for the first time properly define trust propagation and combination in arbitrary OSN topologies by proposing 3VSL (Three-Valued Subjective Logic). The 3VSL distinguishes the posteriori and priori uncertainties existing in trust, and the difference between distorting and original opinions, thus be able to compute multi-hop trusts in arbitrary graphs. We theoretically proved the capability based on the Dirichlet distribution. Furthermore, an online survey system is implemented to collect interpersonal trust data and validate the correctness and accuracy of 3VSL in real world. Both experimental and numerical results show that 3VSL is accurate in computing interpersonal trust in OSNs.

Event-driven data aggregation (EDDA) scenario is widely applied in wireless sensor network (WSN) applications. In this paper, we present a MAC protocol of WSN for EDDA scenario, referred to as E-MAC. Different from other MAC protocols, E-MAC is designed upon the attributes of EDDA scenario. E-MAC classifies the nodes in EDDA scenario and divides the transportation of data packets into several stages. Through sequential wake-up, multi-hop reservation and staggering transmission, E-MAC schedules data packets from the event zone across multiple hops in the same cycle. As a CSMA based protocol, it offers favorable performance with a low energy cost in EDDA scenario. We analyzed the latency of E-MAC and RMAC [3] numerically in EDDA scenario. Besides, we evaluated E-MAC in EDDA scenario using ns-2.29 and compared it with RMAC. Results of both the analysis and the simulation state that the advantage lies in E-MAC in EDDA scenario.