The substantial heterogeneity in energy and communication protocols among IoT devices leads to highly diversified image resolutions, which severely constrain the reliability of downstream visual analysis tasks in applications like intelligent transportation and smart buildings. While recent deep learning-based SR methods have achieved remarkable success, the majority are typically designed for specific integer scaling factors, requiring separate models for different scales, which is impractical for real-world IoT applications. To address this, we propose SafaSR, an arbitrary-scale image super-resolution network based on multi-domain feature fusion. Our key innovation lies in a Multi-domain Multi-level Feature Fusion(M2F2 )mechanism, which is driven by a scale-aware feature learning (SFL) model that adaptively extracts features from both spatial and frequency domains. TheM2F2mechanism is designed to reduce correlations between different feature domains, allowing a more effective integration of complementary information. Extensive experiments show that our proposed network outperforms the most advanced image SR algorithms in terms of PSNR and SSIM metrics on the benchmark datasets, with fewer network parameters and less runtime.

Functional brain connectivity networks capture complex relationships and temporal evolution between brain regions, which have become increasingly important for diagnosing neurological disorders. However, existing methods, which are primarily based on vector or graph representations, struggle to adequately characterize the intricate spatio-temporal topological architecture of functional brain networks. Additionally, they predominantly rely on data-driven paradigms and lack priors pertaining to cross-windows network interactions. To address these issues, we propose a spatio-temporal hypergraph attention network framework for brain network analysis. Specifically, we first propose a temporal attention network architecture embedded with temporal similarity-driven prior knowledge, which effectively extracts long-range dependency information from fMRI by combining multi-head self-attention mechanisms and cross-window temporal prior knowledge. Second, we design a hierarchical hypergraph generation module that fuses local and global brain topological information to achieve multi-scale modeling of high-order spatio-temporal structures. Additionally, the spatial attention network, developed based on transformer architecture, employs hypergraph message passing mechanisms to effectively construct multi-level spatial interaction relationships between brain regions. Finally, a multi-layer perceptron (MLP) is adopted for classification. Experiments on the ADNI and PD datasets demonstrate that our method outperforms several state-of-the-art approaches in diagnostic performance and provides discriminative graph features for relevant brain disease diagnosis.

Recent studies have shown that cooperative multi-agent deep reinforcement learning (c-MADRL) is under the threat of backdoor attacks. Once a backdoor trigger is observed, it will perform abnormal actions leading to failures or malicious goals. However, existing proposed backdoors suffer from several issues, e.g., fixed visual trigger patterns lack stealthiness, the backdoor is trained or activated by an additional network, or all agents are backdoored. To this end, in this paper, we propose a novel backdoor attack against c-MADRL, which attacks the entire multi-agent team by embedding the backdoor only in a single agent. Firstly, we introduce adversary spatiotemporal behavior patterns as the backdoor trigger rather than manual-injected fixed visual patterns or instant status and control the attack duration. This method can guarantee the stealthiness and practicality of injected backdoors. Secondly, we hack the original reward function of the backdoored agent via reward reverse and unilateral guidance during training to ensure its adverse influence on the entire team. We evaluate our backdoor attacks on two classic c-MADRL algorithms VDN and QMIX, in a popular c-MADRL environment SMAC. The experimental results demonstrate that our backdoor attacks are able to reach a high attack success rate (91.6%) while maintaining a low clean performance variance rate (3.7%).

Behavior-oriented decision-makers are critical components in generating intelligent decisions for user virtual interactions in metaverse. In this work, we study the efficiency and security of behavior-oriented decision-makers in metaverse from perspective of autonomous driving (AD), where modeling human uncertain driving behaviors is the key factor of their performance. We first explore the ability of different deep-neural-network-based decision-makers used in deep reinforcement learning for efficient autonomous vehicle control, and then we propose a novel neural backdoor attack against them using spatiotemporal driving behaviors, rather than an immediate state. With our attack, the adversary acts as a normal driver and can trigger attacks by driving his vehicle following specific spatiotemporal behaviors. Extensive experiments show that our proposed backdoor attack can achieve high stealthiness and effectiveness (less than 1% clean performance variance rate and more than 98% attack success rate) on behavior-oriented decision-makers, and is sustainable against existing advanced defenses.

Image super-resolution (SR) is a technique to recover lost high-frequency information in low-resolution (LR) images. Since spatial-domain information has been widely exploited, there is a new trend to involve frequency-domain information in SR tasks. Besides, image SR is typically application-oriented and various computer vision tasks call for image arbitrary magnification. Therefore, in this article, we study image features in the frequency domain to design a novel image arbitrary-scale SR network. First, we statistically analyze LR-HR image pairs of several datasets under different scale factors and find that the high-frequency spectra of different images under different scale factors suffer from different degrees of degradation, but the valid low-frequency spectra tend to be retained within a certain distribution range. Then, based on this finding, we devise an adaptive scale-aware feature division mechanism using deep reinforcement learning, which can accurately and adaptively divide the frequency spectrum into the low-frequency part to be retained and the high-frequency one to be recovered. Finally, we design a scale-aware feature recovery module to capture and fuse multi-level features for reconstructing the high-frequency spectrum at arbitrary scale factors. Extensive experiments on public datasets show the superiority of our method compared with state-of-the-art methods.

Deep reinforcement learning (DRL) has made sig-nificant achievements in many real-world applications. But these real-world applications typically can only provide partial ob-servations for making decisions due to occlusions and noisy sensors. However, partial state observability can be used to hide malicious behaviors for backdoors. In this paper, we explore the sequential nature of DRL and propose a novel temporal-pattern backdoor attack to DRL, whose trigger is a set of temporal constraints on a sequence of observations rather than a single observation, and effect can be kept in a controllable duration rather than in the instant. We validate our proposed backdoor attack to a typical job scheduling task in cloud computing. Numerous experimental results show that our backdoor can achieve excellent effectiveness, stealthiness, and sustainability. Our backdoor's average clean data accuracy and attack success rate can reach 97.8% and 97.5%, respectively.

While providing significant convenience for people, mobile applications (Apps) bring serious privacy leakage and invasion threats over certain platforms (e.g., Android) due to privacy violations. To protect users from these threats, a lot of works related to privacy violation detection have been proposed. However, few of them particularly check the violations, including lacking privacy policy, collecting privacy before statement, lacking account cancelation service, and stubborn permission request. Toward this end, we design an automatic detection tool namedPVDetectorto detect these violations in Android Apps. We extract and construct relevant threat forms by statically and dynamically analyzing Apps’ behaviors, and then fine tune these forms through threat-form-matching methods on problematic Apps. Finally, a comprehensive experiment is conducted to detect privacy violations on different Android application markets byPVDetector. Specifically, we detect 16 162 Android Apps (involving people’s various aspects of life) collected from six popular official application markets and three special categories. The experiment results indicate that the situation that Apps contain privacy violations is greatly serious in these markets and categories. We also randomly check the experiment results of 385 Apps. The check results illustrate that the detection accuracy ofPVDetectorcan reach 93%.

By providing loosely coupled, lightweight, and independent services, the microservice architecture is promising for large-scale and complex service provision requirements in the Internet of Things (IoT). However, it requires more fine-grained resource management and orchestration for service provision. Most of the existing microservice orchestration solutions are based on those designed for the traditional cloud. They can only provide coarse-grained resource allocation using possibly conflicting weighted objectives. In this article, we present a fine-grained microservice orchestration approach to provide services online for dynamic requests of IoT applications. By using a fine-grained resource model of energy cost and service end-to-end response time of orchestrated microservices, we formulate the microservice orchestration problem as a multiobjective Markov decision process. We then propose a multiobjective optimization solution based on deep reinforcement learning (DRL) to simultaneously reduce energy consumption and response time. Through extensive experiments, our proposed algorithm presents significant performance results than the state of the art. To the best of our knowledge, this is the first work that addresses microservice orchestration using DRL for multiple conflicting objectives.

The Internet of Things (IoT) provides convenience for our daily lives via a huge number of devices. However, due to low-resource and poor computing capability, these devices have a high number of firmware vulnerabilities. Software verification is a powerful solution to ensure the correctness and security of IoT firmware programs. Unfortunately, due to the complex semantics and syntax of program languages (typically C), applying software verification in IoT firmware faces the tradeoff between efficiency and accuracy. One of the fundamental reasons is that verification methods cannot support verifying state transitions on the memory space caused by pointer operations well. To this end, by combining sparse value flow (SVF) analysis into model checking and optimizing computational redundancy among them, we design a novel points-to-sensitive model checker, called PCHECKER, which can provide a highly precise and efficient verification for IoT firmware programs. We first design a spatial flow model to effectively describe state behaviors of a C program both on the symbolic and memory space. We then propose a counterexample-guided model checking algorithm that can dynamically refine abstract precisions and update nondeterministic points-to relations. With a set of C benchmarks containing a variety of pointer operations and other complex C features, our experiments have shown that compared with state of the art (SOTA), PCHECKER can achieve outstanding results in the verification tasks of C programs that its verification accuracy is 95.9%, and its average verification time of each line of code is 1.27 ms, which are both better than existing model checkers.

Trigger-action programming (TAP) is a popular end-user programming framework that can simplify the Internet of Things (IoT) automation with simple trigger-action rules. However, it also introduces new security and safety threats. A lot of advanced techniques have been proposed to address this problem. Rigorously reasoning about the security of a TAP-based IoT system requires a well-defined model and verification method both against rule semantics and physical-world features, e.g.,concurrency, rule latency, extended action, tardy attributes,andconnection-based rule interactions, which has been missing until now. By analyzing these features, we find 9 new types of rule interaction vulnerabilities and validate them on two commercial IoT platforms. We then present TAPInspector, a novel system to detect these interaction vulnerabilities in concurrent TAP-based IoT systems. It automatically extracts TAP rules from IoT apps, translates them into a hybrid model by model slicing and state compression, and performs semantic analysis and model checking with various safety and liveness properties. Our experiments corroborate that TAPInspector is practical: it identifies 533 violations related to rule interaction from 1108 real-world market IoT apps and is at least 60000 times faster than the baseline without optimization.

Availability and security problems in cellular emergency call systems can cost people their lives, yet this topic has not been thoroughly researched. Based on our proposed Seed-Assisted Specification method, we start to investigate this topic by looking closely into one emergency call failure case in China. Using what we learned from the case as prior knowledge, we build a formal model of emergency call systems with proper granularity. By running model checking, four public-unaware scenarios where emergency calls cannot be correctly routed are discovered. Additionally, we extract configurations of two major U.S. carriers and incorporate them as model constraints into the model. Based on the augmented model, we find two new attacks leveraging the privileges of emergency calls. Finally, we present a solution with marginal overhead to resolve issues we can foresee.