Environment preservation has become a prominent issue around the world. As traditional internal combustion engine (ICE) vehicles have been major contributors of air pollution, electric (EVs) are gaining popularity. However, due to limited electricity supply battery pack, EVs need be charged frequently and each charge takes long time. This may degrade travel efficiency driver comfort. To address this issue, paper aims minimize charging waiting time through intelligently scheduling activities...

• Drought induces root death in plants; however, the nature and characteristics of cell its underlying mechanisms are poorly understood. Here, we provide a systematic analysis primary tips Arabidopsis during water stress. Root tip occurs when high deficit is reached. The dying cells were first detected apical meristem roots underwent active programmed (PCD). Transmission electron microscopic shows that undergoing induced had unambiguous morphological features autophagic death, including an...

When a sensor network is deployed in hostile environments, the adversary may compromise some nodes, and use compromised nodes to inject false sensing reports or modify sent by other nodes. In order defend against attacks with low cost, researchers have proposed symmetric group key-based en-route filtering schemes, such as SEF [F. Ye et al., March 2004] I-LHAP [S. Zhu 2004]. However, if has large number of many keys can be captured, schemes become ineffective even useless. To deal node...

The emerging wireless charging technology creates a controllable and perpetual energy source to provide power over distance. Schemes have been proposed make use of prolong the sensor network lifetime. Unfortunately, existing schemes only passively replenish sensors that are deficient in supply, cannot fully leverage strengths this technology. To address limitation, we propose J-RoC - practical efficient Joint Routing Charging scheme. Through proactively guiding routing activities delivering...

As wireless power charging technology emerges, some basic principles in sensor network design are changed accordingly. Existing node deployment and data routing strategies cannot exploit to minimize overall energy consumption. Hence, this paper, we (a) investigate the impact of on arrangement, (b) formalize problem, (c) prove it as NP-complete, (d) develop heuristic algorithms solve (e) evaluate performance solutions through extensive simulations. To best our knowledge, is first effort...

Location information is generally considered to be indispensable in wireless sensor networks, but existing algorithms have notable limitations their positioning accuracy 3-dimensional (3D) spaces. To improve the of nodes a 3D scenario, this paper proposes new method called N2-3DDV-Hop (non-dominated sorting genetic algorithm II with distance-vector hop) that builds on 3D-DV-Hop by adding multi-objective model and NSGA-II. In paper, it analyzed traditional single-objective showed relationship...

With accelerated ensemble of the Internet Things technology and automotive industry, vehicular network has been established as powerful tools. However, it is a significant challenge for dynamic heterogeneous to meet high requirements sixth-generation (6G) such reliability security. To address this challenge, we design novel weight-based machine learning algorithm (WBELA) identify abnormal messages Controller Area Network (CAN) bus network. Then, establish model based on many-objective...

Domain generalizable person re-identification (DG ReID) is a challenging problem, because the trained model often not to unseen target domains with different distribution from source training domains. Data augmentation has been verified be beneficial for better exploiting data improve generalization. However, existing approaches primarily rely on pixel-level image generation that requires designing and an extra network, which extremely complex provides limited diversity of augmented data. In...

When deployed for long-term tasks, the energy required to support sensor nodes' activities is far more than that can be preloaded in their batteries. No matter how battery conserved, once used up, network life terminates. Therefore, guaranteeing supply has persisted as a big challenge. To address this problem, we propose node reclamation and replacement (NRR) strategy, with which mobile robot or human labor called repairman (MR) periodically traverses network, reclaims nodes low no power...

Mobile sinks are needed in many sensor network applications for efficient data collection, querying, localized reprogramming, identifying and revoking compromised sensors, other maintenance. Employing mobile however raises a new security challenge: if sink is given too privileges, it will become very attractive attack compromise. Using sink, an adversary may easily bring down or even take over the network. Thus, mechanisms that can tolerate compromises essential. In this paper, based on...

The rapid progress of artificial intelligence expands its wide applicability in Internet Things (IoT). Meanwhile, data insufficient and source privacy are key supply chain challenges facing IoT especially the healthcare industry. To address this problem IoT, article, we propose a skin cancer detection model based on federated learning integrated with deep generation model. First, employ dual generative adversarial networks to data. In addition, improve quality generated images, synchronously...

Unified multimodal large language models (U-MLLMs) have demonstrated impressive performance in visual understanding and generation an end-to-end pipeline. Compared with generation-only (e.g., Stable Diffusion), U-MLLMs may raise new questions about bias their outputs, which can be affected by unified capabilities. This gap is particularly concerning given the under-explored risk of propagating harmful stereotypes. In this paper, we benchmark latest find that most exhibit significant...

In this paper, we propose a bioinspired model for human action recognition through modeling neural mechanisms of information processing in two visual cortical areas: the primary cortex (V1) and middle temporal (MT) dedicated to motion. This model, named V1-MT, is composed V1 MT models (layers) corresponding their areas, which are built with layered spiking networks (SNNs). Some neuron properties MT, such as direction speed selectivity, spatiotemporal inseparability, center surround...