The development of smart homes has advanced the concept user authentication to not only protecting privacy but also facilitating personalized services users. Along this direction, we propose integrate with human-computer interactions between users and household appliances through widely-deployed WiFi infrastructures, which is non-intrusive device-free. In paper, <inline-formula><tex-math notation="LaTeX">$FingerPass$</tex-math></inline-formula> leverages channel state information (CSI)...

Distributed Denial of Service (DDoS) attack is a critical threat to the Internet, and botnets are usually engines behind them. Sophisticated botmasters attempt disable detectors by mimicking traffic patterns flash crowds. This poses challenge those who defend against DDoS attacks. In our deep study size organization current botnets, we found that flows more similar each other compared Based on this, proposed discrimination algorithm using flow correlation coefficient as similarity metric...

This paper advocates a new paradigm of transportation systems for future smart cities, namely, public vehicles (PVs), that provides dynamic ridesharing trips at requests. Passengers will enjoy more convenient and flexible services with much less expense. In the PV system, both number required parking spaces be significantly reduced. There traffic congestion, energy consumption, pollution. this paper, concept, method, algorithm system are described. The key issue effectively implementing is...

Multipath routing can significantly improve the network throughput and end-to-end (e2e) delay. Network coding based multipath removes complicated coordination among multiple paths so that it further enhances data transmission efficiency. Traditional protocols, however, are inefficient for Low Earth Orbit (LEO) satellite networks with long link delay regular topology. Considering these characteristics, in this paper, we first formulate cooperative problem, then propose a Coding Cooperative...

Image-level weakly supervised semantic segmentation has received increasing attention due to its low annotation cost. Existing methods mainly rely on Class Activation Mapping (CAM) obtain pseudo-labels for training models. In this work, we are the first demonstrate that long-tailed distribution in data can cause CAM calculated through classifier weights over-activated head classes and under-activated tail shared features among head- tail- classes. This degrades pseudo-label quality further...

The topology of a vehicular ad hoc network (VANET) changes rapidly due to high-speed movement vehicles, so traditional mobile (MANET) broadcast protocol may not work efficiently in VANET. This paper proposes distance-based called Efficient Directional Broadcast (EDB) for VANET using directional antennas. In EDB, only the furthest receiver is responsible forward packet opposite direction where arrives. Besides, repeater located at intersection helps disseminating packets vehicles on other...

In data center networks, how to balance workloads is a key issue with the fast growth of network applications. Open Flow protocol, which competitive candidate for solving problem, provides each user programmatic control specific flows, so as determine their paths through network. However, existing solutions based on only try find static routing path during initialization step while often suffers from poor performance since configuration may change transmission. To solve this paper proposes...

Load-balanced flow scheduling for big data centers in clouds, which a large amount of needs to be transferred frequently among thousands interconnected servers, is key and challenging issue. The OpenFlow promising solution balance flows center networks through its programmatic traffic controller. Existing based schemes, however, statically set up routes only at the initialization stage transmissions, suffers from dynamical distribution changing network states often results poor system...

Software-defined networking (SDN) based LEO satellite networks can make full use of resources through flexible function configuration and efficient resource management controllers. Consequently, controllers have to be carefully deployed on dynamical topology time-varying workload. However, existing work controller placement assignment is not applicable with highly dynamic randomly fluctuating load. In this paper, we first formulate the adaptive (ACPA) problem prove its NP-hardness. Then,...

Graph-based social recommendation improves the prediction accuracy of by leveraging high-order neighboring information contained in relations. However, most them ignore problem that relations can be noisy for recommendation. Several studies attempt to tackle this performing graph denoising, but they suffer from 1) adaptability issues other graph-based models and 2) insufficiency user representation learning. To address limitations, we propose a model-agnostic denoising module (denoted as...

SDN (Software defined networking) provides effective technical methods for optimal resource management. However, there are conflicts frequent and serious in current related schemes because they mix elephant mice flows on shared transmission paths. So, controllers have to be smart enough detect with low cost then reroute a feature-aware way. existing flow detection suffer from high bandwidth consumption long time; little literature considers mice-flow scheduling. In this paper, we propose an...

It is a desirable goal to fuse satellite and ground integrated networks (SGINs) improve the resource utilization efficiency. However, existing work did not consider how integrate them as whole network because they lack of function-configurable management efficient cooperation among networks. In this paper, we firstly propose SDN-based architecture, where resources in SGINs are managed scheduled layered on-demand way. Then, formulate dynamical transmission an optimization problem prove its NP...

The continuously increasing time spent on car trips in recent years brings growing attention to the physical and mental health of drivers roads. As one key vital signs, heartbeat is a critical indicator drivers' states. Most existing studies monitoring either require sensor attachment or could only provide sketchy heart rates. Moreover, most approaches subject remain stationary quiet measuring environment, which hard apply dynamic driving environments. In this paper, we propose contactless...

Information networks generally exhibit three characteristics, namely dynamicity, heterogeneity, and node attribute diversity. However, most existing network embedding approaches only consider two of the when each into low-dimensional space. Adding to such an approach a technique processing remaining characteristic can easily cause incompatibility. One solution process characteristics together is treat dynamic heterogeneous attributed (DHAN) as temporal sequence (HAN) snapshots. For example,...

Cognitive radio significantly mitigates the spectrum scarcity for various applications built on wireless communication. Current techniques mobile cognitive ad hoc networks (MCADNs), however, cannot be directly applied to time-critical due channel interference, node mobility as well unexpected primary user activities. In multichannel multiflow MCADNs, it becomes even worse because multiple links potentially interfere with each other. this paper, we propose a delay-minimized routing (DMR)...

Recent years have witnessed the emerging market of 3D facial reconstruction that supports numerous face-driven scenarios including modeling in virtual reality (VR), human-computer interaction, and affective computing applications. Current mainstream approaches rely on vision for reconstruction, which may encounter privacy concerns suffer from obstruction scenes bad lighting conditions. In this paper, we present a nonintrusive system, mm3DFace, leverages millimeter wave (mmWave) radar to...

Image segmentation plays an important role in vision understanding. Recently, the emerging foundation models continuously achieved superior performance on various tasks. Following such success, this paper, we prove that Segment Anything Model 2 (SAM2) can be a strong encoder for U-shaped models. We propose simple but effective framework, termed SAM2-UNet, versatile image segmentation. Specifically, SAM2-UNet adopts Hiera backbone of SAM2 as encoder, while decoder uses classic design....

Long-sequence causal reasoning seeks to uncover relationships within extended time series data but is hindered by complex dependencies and the challenges of validating links. To address limitations large-scale language models (e.g., GPT-4) in capturing intricate emotional causality dialogues, we propose CauseMotion, a long-sequence framework grounded Retrieval-Augmented Generation (RAG) multimodal fusion. Unlike conventional methods relying only on textual information, CauseMotion enriches...