Surface defect detection plays an important role in particleboard quality control. But it still faces challenges detecting coexisting multi-scale defects and weak texture defects. To address these issues, this study proposed PBD-YOLO (Particleboard Defect-You Only Look Once), a lightweight YOLO-based algorithm with feature fusion enhancement capabilities. In order to improve the ability of extract features, SPDDEConv (Space Depth Difference Enhance Convolution) module was introduced study,...

Physical-layer key generation has emerged as a promising avenue for establishing secret keys using reciprocal channel measurements between wireless devices. However, reciprocity may suffer degradation from ambient noise and cause mismatched bits, while existing methods mitigating this issue yet face limitations in efficiency. The root behind such is the heavy reliance on measurements, which can be naturally susceptible to non-reciprocity attributed environmental factors. Instead of direct...

Abstract Combinatorial optimization problems have very important applications in information technology, transportation, economics, management, network communication, and other fields. Since the problem size real-scenario application is large-scale, demand for real-time efficient solving approaches increases rapidly. The traditional exact methods guarantee optimality of final solution, but these can hardly solve acceptable time due to extremely high computational costs. Heuristic find...

The emerging smart city is driving massive transformations of modern cities, facing the huge influx sensor data from IoT devices. Edge computing distributes tasks to near-edge end, which greatly enhances service quality applications, that is, ultralow latency, large capacity, and high throughput. However, due constrained resource devices, currently, systems with a centralized model are vulnerable attacks, such as DDoS botnet central database failure, can hardly provide high-confidence...

Personal Identification Number (PIN), as one of the primary means protecting digital properties and privacy on mobile devices, has been suffering from shoulder surfing attacks weak password guessing for long term. Recent years witness growing interest in two-factor authentication that takes advantage two different ways mutual verification, thereby strengthening user authentication's accuracy reliability. Especially with popularity smartwatches, more physiological signals are readily...

Summary Electrical fires caused by leakage, short circuit, overload, and other electrical failures occur frequently in residential areas, failure to find locate the fault location time is main reason for evolution of circuit into fire, especially ancient buildings with aging facilities. Different from modern large‐scale buildings, architectural complex composed a large number scattered low inconsistent facility specifications simple structure, so it does not have conditions deploy unified...

A photovoltaic (PV) plant is normally built in a fixed series-parallel configuration and its power-voltage characteristics often get complex with multiple peaks under partial shading scenarios. Therefore, identification of important for monitoring invoking maximum power pint estimation. This paper proposes back-propagation neural network (BPNN) based method which locates shaded modules by using measured voltage data. Optimal sensor placement schemes are introduced to decrease the number...

The physical layer secret key generation exploiting wireless channel reciprocity has demonstrated its viability and effectiveness in various scenarios, such as the Internet of Things (IoT) network, mobile communication industrial control system. Most existing studies rely on quantization technique to convert measurements into bits for confidential communications. However, non-simultaneous packet exchanges time-division duplex systems noise effects usually induce inconsistent results...

As one of the most common ways for user authentication, Personal Identification Number (PIN), due to its simplicity and convenience, has suffered from plenty side-channel attacks, which pose a severe threat people’s privacy property. The success existing attacks is usually built upon premise no occlusion between attacker victim’s hand gesture, but it increases difficulty launching attack possibility exposure. To overcome such limitation, we propose ArmSpy++, an improved video-assisted PIN...

This paper opens up a new pathway of the utility breathing pattern for user verification.We demonstrate that it is possible to capture people's leveraging commodity WiFi devices.While prior solutions biometricsbased recognition usually require dedicated devices (e.g., video cameras or IR sensors), this introduces first general, low-cost breathing-based verification system using devices.The proposed based on fact always keeps consistent same but distinct among different people.Our innovative...

PIN inference attack leveraging keystroke-induced side-channel information poses a substantial threat to the security of people's privacy and properties. Among various attacks, video-assisted method provide more intuitive robust infer PINs. But it usually requires there is no visual occlusion between attacker victims or their hand gestures, making attackers either easy expose themselves inapplicable scenarios such as ATM POS terminals. In this paper, we present novel practical system,...

The popularity of intelligent mobile devices enables users to access various information services anytime and anywhere, which increases people's demand for communication networks. WiFi network is one the main ways smart network. Compared with using cellular network, are more willing connect when there coverage, even automatically crack password connect. In order improve security devices, this paper proposes an Authentication method connection device based on Channel State Information (CSI)....

In distributed antenna systems (DAS), multi-user interference was shown to be asynchronous, which had not been paid much attention. this study, the authors compare performances of DAS under condition asynchronous (AI) with four well-known multiple-input multiple-output (MIMO) algorithms namely zero-forcing (ZF), block diagonalisation (BD), signal-to-leakage-plus-noise ratio (SLNR) and dirty paper coding (DPC) algorithms. The simulation results clearly demonstrate significant performance...

Similarity search for trajectories, especially the top-k similarity query, has been widely used in different fields, such as personalized travel route recommendation, car pooling, etc. Previous works have studied trajectory query plaintext, but increasing attention to privacy protection makes on data become a challenge. In this paper, we propose privacy-preserving scheme over large-scale based Hilbert curve and homomorphic encryption. Towards end, first define spatio-temporal measure that...

The interest of indoor movement state recognition has been recently extensively evaluated by several research teams. This paper fully explores the signals from all sensors in a smartphone and develops novel deep neural network-based model. In off-line training phase, this firstly proposes matrix to pre-processing raw data overcome its redundancy complexity. second place, managed go through stacked denoising autoencoder network, resulting an output layer with seven different states. trained...