Efficiency Of Computing Resources Research Articles

A Sustainable W-RLG Model for Attack Detection in Healthcare IoT Systems

The increasingly widespread use of IoT devices in healthcare systems has heightened the need for sustainable and efficient cybersecurity measures. In this paper, we introduce the W-RLG Model, a novel deep learning approach that combines Whale Optimization with Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Units (GRU) for attack detection in healthcare IoT systems. Leveraging the strengths of these algorithms, the W-RLG Model identifies potential cyber threats with remarkable accuracy, protecting the integrity and privacy of sensitive health data. This model’s precision, recall, and F1-score are unparalleled, being significantly better than those achieved using traditional machine learning methods, and its sustainable design addresses the growing concerns regarding computational resource efficiency, making it a pioneering solution for shielding digital health ecosystems from evolving cyber threats.

Deep Reinforcement Learning-Empowered Cost-Effective Federated Video Surveillance Management Framework.

Video surveillance systems are integral to bolstering safety and security across multiple settings. With the advent of deep learning (DL), a specialization within machine learning (ML), these systems have been significantly augmented to facilitate DL-based video surveillance services with notable precision. Nevertheless, DL-based video surveillance services, which necessitate the tracking of object movement and motion tracking (e.g., to identify unusual object behaviors), can demand a significant portion of computational and memory resources. This includes utilizing GPU computing power for model inference and allocating GPU memory for model loading. To tackle the computational demands inherent in DL-based video surveillance, this study introduces a novel video surveillance management system designed to optimize operational efficiency. At its core, the system is built on a two-tiered edge computing architecture (i.e., client and server through socket transmission). In this architecture, the primary edge (i.e., client side) handles the initial processing tasks, such as object detection, and is connected via a Universal Serial Bus (USB) cable to the Closed-Circuit Television (CCTV) camera, directly at the source of the video feed. This immediate processing reduces the latency of data transfer by detecting objects in real time. Meanwhile, the secondary edge (i.e., server side) plays a vital role by hosting a dynamically controlling threshold module targeted at releasing DL-based models, reducing needless GPU usage. This module is a novel addition that dynamically adjusts the threshold time value required to release DL models. By dynamically optimizing this threshold, the system can effectively manage GPU usage, ensuring resources are allocated efficiently. Moreover, we utilize federated learning (FL) to streamline the training of a Long Short-Term Memory (LSTM) network for predicting imminent object appearances by amalgamating data from diverse camera sources while ensuring data privacy and optimized resource allocation. Furthermore, in contrast to the static threshold values or moving average techniques used in previous approaches for the controlling threshold module, we employ a Deep Q-Network (DQN) methodology to manage threshold values dynamically. This approach efficiently balances the trade-off between GPU memory conservation and the reloading latency of the DL model, which is enabled by incorporating LSTM-derived predictions as inputs to determine the optimal timing for releasing the DL model. The results highlight the potential of our approach to significantly improve the efficiency and effective usage of computational resources in video surveillance systems, opening the door to enhanced security in various domains.

Deep-Learning-Based Action and Trajectory Analysis for Museum Security Videos

Recent advancements in deep learning and video analysis, combined with the efficiency of contemporary computational resources, have catalyzed the development of advanced real-time computational systems, significantly impacting various fields. This paper introduces a cutting-edge video analysis framework that was specifically designed to bolster security in museum environments. We elaborate on the proposed framework, which was evaluated and integrated into a real-time video analysis pipeline. Our research primarily focused on two innovative approaches: action recognition for identifying potential threats at the individual level and trajectory extraction for monitoring museum visitor movements, serving the dual purposes of security and visitor flow analysis. These approaches leverage a synergistic blend of deep learning models, particularly CNNs, and traditional computer vision techniques. Our experimental findings affirmed the high efficacy of our action recognition model in accurately distinguishing between normal and suspicious behaviors within video feeds. Moreover, our trajectory extraction method demonstrated commendable precision in tracking and analyzing visitor movements. The integration of deep learning techniques not only enhances the capability for automatic detection of malevolent actions but also establishes the trajectory extraction process as a robust and adaptable tool for various analytical endeavors beyond mere security applications.

Tracking scheme of airborne star tracker based on event-triggered sliding mode control with known input delay

The airborne star tracker is crucial in aircraft navigation systems, with its tracking performance directly impacting navigation accuracy. Under airborne conditions, the performance of its tracking control will be compromised by various disturbances. Moreover, the limitation in computational resources is another issue that must be addressed. Assuming that the existing studies on this application did not consider these two aspects of the effects simultaneously, this study proposes a novel event-triggered sliding mode control (ET-SMC) scheme considering the known input time delay for star tracking control to address these two issues. First, an extended state observer (ESO) is presented to estimate the disturbance generated by airborne conditions. Second, the ET-SMC scheme further enhances the robustness and improves resource utilization, thus easing the processor burden. An ET mechanism related to the disturbance estimation triggering error in the ESO is introduced to ensure that the system input is only updated when necessary. A novel, easy-to-implement sliding gain is also proposed to increase system adaptability and reduce inherent chattering. The reachability of the sliding surface and the existence of a practical sliding mode of the system are ensured based on the Lyapunov theory. The ultimate upper bound of system states considering the known input time delay is also proven, thereby confirming the stability of the proposed design. The exclusion of Zeno behavior validates the feasibility of the proposed ESO-based ET-SMC. Finally, the effectiveness of the proposed method is verified using comparative simulations and target-tracking experiments. The experimental results demonstrate that the proposed method excels in robustness, disturbance attenuation, high tracking accuracy, and computational resource efficiency. These enhancements are anticipated to result in a more stable tracking performance for the star tracker, thereby contributing to precise aircraft navigation.

Enhancing data privacy in wireless sensor network using homomorphic encryption

From environmental monitoring to healthcare, WSN are essential. The extensive use of these networks to gather sensitive data requires strong data privacy mechanisms. Cryptographic homomorphic encryption may solve this problem. This study uses CKKS encryption and ciphertext packing to improve Wireless Sensor Network data privacy. CKKS is a fully homomorphic encryption scheme that allows complex data operations. This makes it ideal for data analytics and computations. Ciphertext packing improves computational efficiency, making this combination a good choice for data protection in WSN environments with limited resources. Performance efficiency is crucial to evaluation. This parameter evaluates CKKS encryption with ciphertext packing’s computational resource and execution time efficiency. This study examines how this approach affects WSN node and network performance, revealing its feasibility and practicality in real-world situations. The Data Security parameter highlights CKKS encryption with ciphertext packing’s data security. This comprehensive assessment determines data confidentiality and integrity in the Wireless Sensor Network. The system’s ability to withstand data breaches and unauthorized access is crucial for protecting sensitive data. Any WSN solution must consider energy consumption. This parameter evaluates data encryption and processing energy usage, taking sensor node energy constraints into account. The assessment of CKKS encryption with ciphertext packing in WSN sheds light on its potential to improve data confidentiality while addressing WSN challenges. This research strengthens WSN data protection efforts and lays the groundwork for future advances in this vital field.

75MW AC PV Module Field Anomaly Detection Using Drone-Based IR Orthogonal Images With Res-CNN3 Detector

In recent years, the use of solar photovoltaic (PV) technologies for generating electricity has gained much popularity due to their efficiency, cost-effectiveness, reliability and the need to reduce carbon emissions and air pollution levels around the world in order to control and limit global climate change. Drone-based inspection of Solar Plants is an efficient method to perform preventative and corrective maintenance on the Solar PV arrays installed in large-scale grid-connected Solar PV Plants. Quite a few studies have been made in detecting PV module anomalies from these drone-captured thermal images. The main challenge that most of the applied techniques and models in these studies experience is the capability to detect and localize various PV module defects with high and consistent confidence scores and accuracy. The efficiency and robustness of anomaly detection of these models are also reduced when the testing image scales vary. In this article, we propose the Res-CNN3 framework for defective PV modules region object detection. Res-CNN3 follows a bipartite process for defect regions detection to process the thermal delta of a thermal PV module image using a combination of concatenated CNNs and residual networks to identify whether the thermal image is characteristic of defect regions or not. Logistic regression is employed as the loss function whist the Selective Search (SS) algorithm is utilized to predict the regions of interest (RoI) of the input thermal images for PV module defect regions object detection. Several categories of experiments are performed on the consolidated dataset. The results of these experiments demonstrate that the proposed model surpasses the advanced benchmarked methods in accuracy and computing resource efficiency.

Cheetah: An Accurate Assessment Mechanism and a High-Throughput Acceleration Architecture Oriented Toward Resource Efficiency

Convolutional neural network (CNN) is widely used in artificial intelligence for its excellent recognition accuracy. With its scale increasing rapidly and architecture becoming complicated, it is much difficult to implement CNN in hardware platform efficiently. Many FPGA-based CNN accelerators are proposed in previous work. However, when evaluating resource efficiency, their assessment methods are: 1) device related; 2) frequency related; or 3) they confuse resource efficiency with resource occupancy. There is an insistent demand for intuitive and fair assessment criteria. When implementing CNNs, they still have improvement room in computing resource efficiency, especially for layers with large feature size and few feature maps. In this work, we propose $R_{\mathrm{ score}}$ and $C_{\mathrm{ score}}$ , which compose a comprehensive and accurate resource efficiency assessment mechanism for evaluation and design guidance, respectively. Under the guidance, we introduce Cheetah, an FPGA-based high-throughput acceleration architecture. Its computing part can optimize the use of available resources in both time and space aspects, resulting in better throughput improvement. An auxiliary storage system and a pipeline stage compression method are designed for less storage overhead and shorter inference latency. We implement AlexNet and ResNet18 on KCU1500 at 230 and 240 MHz, respectively, with a throughput of 2411.01GOP/s and 2435.05GOP/s for 16-bit quantification. Cheetah achieves an excellent average $R_{\mathrm{ score}}$ of (0.9441, 0.9456) on different FPGA devices, while the others’ mainly distribute between 0.3 and 0.8. Finally, Cheetah has 6.78X speed improvement and 1.87X power-efficiency improvement than that of Nvidia Jetson TX2, which is the fastest, most power-efficient embedded AI computing device.

Industrial Grade Adaptive Control Scheme for a Micro-Grid Integrated Dual Active Bridge Driven Battery Storage System

In this paper, an industrial grade adaptive control scheme is proposed for a micro-grid integrated dual active bridge driven battery management system (DIBMS). A benchmark industrial grade adaptive control scheme depends on two factors namely robustness and computational resource utilization when such controllers are implemented over processors. The mathematical model of DIBMS system is nonlinear, thus for desired response, non-linear controllers based on sliding mode variable structure control theory suits it well for the state regulation problem of DIBMS, however such controllers utilize high computational resources when practically implemented over processors. Keeping in view the above performance indices, this paper proposes an industrial grade computationally efficient and finite time adaptive robust convergent control for DIBMS system. A proportional integral (PI) scheme is used as central control unit and Hebbian algorithm with double integration of the state error is introduced for online tuning the gains of central control unit. The robustness and computational resource efficiency of the proposed control paradigm is validated using a laboratory scale test bench through TI Launchpad (TMS320F28379D). The superiority of the proposed AI based PI control paradigm is compared with classical PI, integer order sliding mode control (SMC), and fractional order SMC (FOSMC) in terms of computational resource utilization and robustness under all test conditions.

A Fast Compressed Sensing Decoding Technique for Remote ECG Monitoring Systems

Compressed Sensing (CS) has been proposed as a low-complexity ECG data compression scheme for wearable wireless bio-sensor devices. However, CS decoding is characterized by high computational complexity. As a result, it represents a burden to the computational and energy resources of the network gateway node, where decoding is performed. In this article, we propose a Fast Compressive Electrocardiography (FCE) technique to address this problem. CS decoding in FCE is based on Weighted Regularized Least-Squares (WRLS), rather than the standard approach based on ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> norm minimization. The WRLS formulation takes into account prior knowledge of ECG signal properties to estimate an optimally compact and accurate representation of ECG signals. Numerical results show that decoding by FCE is on average 33 times faster than the fastest tested CS-based ECG decoding technique. In addition, high-quality ECG signal reconstruction by FCE is achieved at 32% higher compression ratio. Therefore, FCE can contribute to improving the overall energy and computational resource efficiency of CS-based remote ECG monitoring systems.

A Comparison of Medium-Sized Basis Sets for the Prediction of Geometries, Vibrational Frequencies, Infrared Intensities and Raman Activities for Water

Optimized geometries, vibrational frequencies, as well as infrared intensities and Raman activities were calculated for water (H2O) utilizing popular quantum mechanical approaches. Here, density functional theory (DFT) calculations were performed using the B3LYP (Becke, three-parameter, Lee-Yang-Parr) functional, as well as ab initio calculations using second-order Møller-Plesset (MP2) perturbation theory and coupled-cluster with single, double and perturbative triple excitations [CCSD(T)] levels of theory were used. We assess and benchmark the performance of 69 different atomic orbital basis sets including various popular families of medium-sized basis sets typically of two to four zeta quality and differing levels of augmentation by polar and diffuse functions. The basis sets range from the commonly adopted Pople-style (6-31G & 6-311G), Dunning’s correlation consistent (cc-pV(n+d)Z & aug-cc-pV(n+d)Z, as well as Truhlar’s calendar variations, Jensen’s polarization consistent (pc-n & aug-pc-n), Ahlrichs (def2-…), Sapporo’s and Karlsruches as well as atomic natural orbitals (ANOs) such as NASA Ames (ANOn), Neese-style, and Roos-style. We also compare several basis sets specifically designed to calculate vibrational and electronic properties, including the Sadlej-pVTZ (and LPol-X families), as well as SNS families of Barone. The results are compared to experimental values where available, or calculations performed with 5 or 6 zeta-level (e.g., cc-pV6Z). The performance of each family of basis sets is discussed in terms of their accuracy (and pitfalls), as well as computational resource scaling and efficiency. The Def2 basis family performs very well overall, yielding more accurate results with lower runtimes than traditional basis sets. ‘May’ basis sets also provide accurate predictions of vibrational frequencies at significantly lower costs. Raman activities can be accurately calculated using MP2 under harmonic approximation with several ‘spectroscopic’ families performing well.

Fog Computing Enabled Future Mobile Communication Networks: A Convergence of Communication and Computing

The convergence of communication and computing (COM2P) has been taken as a promising solution for the sustainable development of mobile communication systems. The introduction of fog computing in future mobile networks makes COM2P possible. This article provides an overview on fog computing enabled mobile communication networks (FogMNW), including network architecture, system capacity and resource management. First, this article analyzes the heterogeneity of FogMNW with both advanced communication techniques and fog computing. Then a heterogeneous communication and hierarchical fog computing network architecture is proposed. With both communication and computing resources, FogMNW is enabled to achieve much higher capacity than conventional communication networks. This has been well demonstrated by the coded multicast scheme. Furthermore, a systematic management of communication and computing resources is necessary for FogMNW. By exploiting the communication load diversity in N cells, a communication load aware CLA scheme can achieve much higher computing resource efficiency than comparing schemes. The performance gap increases with N, and CLA can improve efficiency by more than 100 percent when there are 14 cells.

Heterogeneous Ultradense Networks with NOMA: System Architecture, Coordination Framework, and Performance Evaluation

Heterogeneous ultradense networks (H-UDNs) are one key enabler for fifth-generation (5G) wireless networks and beyond to satisfy the explosive growth of mobile data traffic, which exploits spatial reuse of scarce spectrum by deploying massive base stations (BSs) to boost network capacity and enhance network coverage. In this article, we present the system architecture for 5G H-UDNs, consisting of virtualized integrated ground-air-space radio access networks (RANs) and core networks and study network coordination for virtualized H-UDN to efficiently manage computing resources and intercell interference. We look at a cloud-fog-computing coordination framework for efficient computing resource management by achieving reasonable computing task distribution and transfer; computing load balance for computing tasks among virtual computing resources to improve network performance and computing resource efficiency; and a macro-small cell coordination framework for virtualized H-UDN with nonorthogonal multiple access (NOMA) to efficiently manage intercell interference and improve network performance. The joint macro-small enhanced intercell interference coordination (eICIC) and small-small coordinated multipoint joint transmission (CoMP-JT) scheme can efficiently eliminate macro-small intercell interference and utilize small-small intercell interference.

MPI_XSTAR: MPI-based Parallelization of the XSTAR Photoionization Program

We describe a program for the parallel implementation of multiple runs of xstar, a photoionization code that is used to predict the physical properties of an ionized gas from its emission and/or absorption lines. The parallelization program, called mpi_xstar, has been developed and implemented in the C++ language by using the Message Passing Interface (MPI) protocol, a conventional standard of parallel computing. We have benchmarked parallel multiprocessing executions of xstar, using mpi_xstar, against a serial execution of xstar, in terms of the parallelization speedup and the computing resource efficiency. Our experience indicates that the parallel execution runs significantly faster than the serial execution, however, the efficiency in terms of the computing resource usage decreases with increasing the number of processors used in the parallel computing.

Analysis of variants of Cholesky method for solution of big sparse systems of linear algebraic equations

Algorithms of solution of big sparse systems of linear algebraic equations based on Cholesky technique is released in program complex of analysis of stress-strain condition under cyclic elastoplastic loading by finite element method. Comparison of memory and computational resource efficiency of algorithms is carried out.

분산 환경에서 계산 자원의 효율 증대를 위한 데이터 특성 기반의 작업 분류방법

분산 환경에 존재하는 다양한 이기종의 계산 자원은 가상화 기술을 통해 통합된 고성능 컴퓨팅 환경을 구축한다. 최근, 사용자 수준의 향상으로 인해 복잡한 응용 작업의 처리에 대한 요구가 증가하고 있으며, 이는 고성능 컴퓨팅에 대한 수요로 이어지고 있다. 사용자가 요구하는 각각의 작업에는 데이터가 포함되어 있고, 각각의 데이터는 고유의 특성을 가지고 있으므로, 작업의 분류와 처리는 데이터의 특성이 고려되어야 한다. 본 논문에서는 분산 환경에서 계산 자원의 효율 증대를 위한 데이터 특성 기반의 작업 분류방법(JCDT : Job Classifying method based on Data Traits for Increased Efficiency of Computational Resources in Distributed Environment) 을 제안한다. 제안하는 JCDT 는 사용자가 요구하는 작업이 지닌 데이터의 특성을 기반으로 작업을 분류하여, 계산 자원의 효율 증대와 작업 처리속도를 향상시킬 수 있을 것으로 기대한다. Various computational resources in distributed environment are to build a high-performance computing environments through virtualization technology. Recently, there is a growing need for a complicated process due to the improvement of the user-level application, which has led to demand for high-performance computing. The requested job from users is composed of data. And because of each data has own characteristics, the classifier may consider the features of data. In this paper, we propose Job Classifying method based on Data Traits for Increased Efficiency of Computational Resources in Distributed Environment (JCDT). JCDT classifies the job by data traits of the users' request, is expected to improve the job processing time and increase the processing speed of the calculation resources.

Bioinformatics algorithm development for Grid environments

A Grid environment can be viewed as a virtual computing architecture that provides the ability to perform higher throughput computing by taking advantage of many computers geographically dispersed and connected by a network. Bioinformatics applications stand to gain in such a distributed environment in terms of increased availability, reliability and efficiency of computational resources. There is already considerable research in progress toward applying parallel computing techniques on bioinformatics methods, such as multiple sequence alignment, gene expression analysis and phylogenetic studies. In order to cope with the dimensionality issue, most machine learning methods either focus on specific groups of proteins or reduce the size of the original data set and/or the number of attributes involved. Grid computing could potentially provide an alternative solution to this problem, by combining multiple approaches in a seamless way. In this paper we introduce a unifying methodology coupling the strengths of the Grid with the specific needs and constraints of the major bioinformatics approaches. We also present a tool that implements this process and allows researchers to assess the computational needs for a specific task and optimize the allocation of available resources for its efficient completion.

휴대용통신단말의 효과적인 보안관리를 위한 보안 재구성기법의 설계 및 구현

A mobile communication device is a small size of portable computer which provides communication service, such as smart phone and PDA. Currently, one of the biggest barriers in developing the mobile communication device is security issue. Even though there are excellent security functions which can remove the security issues, there is a problem that the mobile communication device can not be loaded with all the functions because it has low storage, poor computational power, and inconvenient user interface, compared to the desktop personal computer. This paper proposes a context-aware security reconfiguration scheme for effective security management of the mobile communication device. The scheme can provide the mobile communication device with the optimized security service which is most adapted to its current security context. Through the prototype implementation and the experiments of the proposed scheme, we have confirmed that the proposed scheme is excellent in terms of computing resource efficiency and usability, without degrading security level.

Flexible Grid service management through resource partitioning

In this paper, a distributed and scalable Grid service management architecture is presented. The proposed architecture is capable of monitoring task submission behaviour and deriving Grid service class characteristics, for use in performing automated computational, storage and network resource-to-service partitioning. This partitioning of Grid resources amongst service classes (each service class is assigned exclusive usage of a distinct subset of the available Grid resources), along with the dynamic deployment of Grid management components dedicated and tuned to the requirements of a particular service class introduces the concept of Virtual Private Grids. We present two distinct algorithmic approaches for the resource partitioning problem, the first based on Divisible Load Theory (DLT) and the second built on Genetic Algorithms (GA). The advantages and drawbacks of each approach are discussed and their performance is evaluated on a sample Grid topology using NSGrid, an ns-2 based Grid simulator. Results show that the use of this Service Management Architecture in combination with the proposed algorithms improves computational and network resource efficiency, simplifies schedule making decisions, reduces the overall complexity of managing the Grid system, and at the same time improves Grid QoS support (with regard to job response times) by automatically assigning Grid resources to the different service classes prior to scheduling.