Expert Systems with Applications, Journal Year: 2024, Volume and Issue: unknown, P. 125879 - 125879
Published: Dec. 1, 2024
Language: Английский
Electronics, Journal Year: 2024, Volume and Issue: 13(5), P. 826 - 826
Published: Feb. 20, 2024
Dynamic voltage and frequency scaling (DVFS) is a technique used to optimize energy consumption in ultra-low-power embedded systems. To ensure sufficient computational capacity, the system must scale up its performance settings. The objective conserve times of reduced demand and/or when battery power used. Fast Fourier Transform (FFT), Cyclic Redundancy Check 32 (CRC32), Secure Hash Algorithm 256 (SHA256), Message-Digest 5 (MD5) are focused functions that achieve energy-efficient performance. Selected operations analyzed from perspective. In this manner, required perform specific function observed, thereby mitigating influence instruction set or architecture. For stable operating scaling, an exponential model for calculation presented. Statistical significance tests conducted validate support findings. Results show proposed optimization reduces applications 27.74% 47.74%.
Language: Английский
Citations
6
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 87 - 96
Published: Jan. 25, 2024
The process of accomplishing strategic objectives by concentrating on effects as opposed to attrition-based destruction is known effects-based operations, or EBO. Finding important nodes in an adversary network a critical step the EBO for successful implementation. In this paper, propose network-based method identify most influential combining centrality and optimization. To determine node influence, adversary's structure analyzed using degree between centralities. Given dynamic nature struct[1]ure results, optimization model that takes resource constraints into account chooses key nodes. Our findings demonstrate various properties, such centralities, influence priorities targets, yields better with decreasing marginal properties. There discussion implications theory sensible decision-making.
Language: Английский
Citations
5
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 68 - 77
Published: Jan. 25, 2024
Cloud computing revolutionizes fast-changing technology. Companies' computational resource use is changing. Businesses can quickly adapt to changing market conditions and operational needs with cloud-based solutions' adaptability, scalability, cost-efficiency. IT operations service delivery have changed due widespread access. efficiently allocates resources in cloud environments, making it crucial this transformation. Resource allocation impacts efficiency, cost, performance, SLAs. Users providers allocate based on workloads using elasticity, on-demand provisioning. economics effectiveness rapid flexible allocation. Proactive versus reactive key understanding management challenges opportunities. Reactive strategies only when shortages or surpluses occur at demand. This responsive strategy often leads inefficiencies like over- under-allocation, which raises costs lowers performance. Predictive analysis workload forecasting predict proactive Optimize avoid over-provisioning. Attention has been drawn predictive These methods historical data, machine learning, analytics. optimize by considering future decisions. Reduced bottlenecks boost user satisfaction lower costs. Matching distribution optimizes management. prediction improves deep learning. CNN, LSTM, Transformer algorithms are promising. New tools for accurate predictions come from their ability spot intricate patterns data. paper compares learning forecasting. study determines the best accuracy ada[1]ptability algorithm Google Cluster Data (GCD). The evaluates upgrading model. advances strategies, help organizations improve utilization, cost-effectiveness, performance face of technological change.
Language: Английский
Citations
3
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 47 - 57
Published: Jan. 25, 2024
For Wireless Sensor Networks (WSNs) to operate as efficiently possible in Indoor Internet of Things (IoT) environments, energy-efficient localization approaches are essential. We investigate several approaches, such trilateration based on Received Signal Strength Indicator (RSSI), Proximity Based Technique, Inertial Navigation, Ultrasound-based, and Magnetic Field-based the context energy efficiency. RSSI-based trilateration, which provides good accuracy with little consumption, uses measurements signal intensity infer device positions. In cases where there limitations line sight, technologies ultrasound measure travel durations. Although calibration sensitivity interference taken into account, magnetic field-based use field anomalies determine Accuracy, usage, scalability, robustness, effort some factors that these techniques evaluated against order fulfil demands indoor IoT environments. A thoughtful choice methods can increase efficiency, lifespan sensor networks, enable precise location-aware applications. meet increasing demand for localisation more research this is still being conducted.
Language: Английский
Citations
2
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 38 - 46
Published: Jan. 25, 2024
The widespread use of Wireless Sensor Networks (WSN) in Internet Things (IoT) causes energy efficiency issues. This paper proposes an AI-based solution to this problem. propose AI-Driven Power Optimization framework for IoT-enabled WSN using Deep Q-Network (DQN) and Dynamic Voltage Frequency Scaling (DVFS). These techniques can adapt changing network conditions reduce power consumption when used together. nodes provide environmental parameters, battery status, behavior data the AI-driven DQN is implemented after preprocessing learn make management decisions reinforcement learning. Neural network-driven agent operates a state action space. It optimizes with rewards. Real-time hardware adjustment done DVFS. DVFS precise control decision-making create comprehensive optimization strategy. AI adapts new challenges lifespan by improving its policies. Experimental implementations proposed show significant savings, extension, QoS improvements. proven effective flexible. study shows potential AI, specifically DVFS, WSN. addresses improves IoT sensor networks. making deployments more sustainable resilient.
Language: Английский
Citations
2
Lecture notes on data engineering and communications technologies, Journal Year: 2024, Volume and Issue: unknown, P. 137 - 163
Published: Jan. 1, 2024
Language: Английский
Citations
2
Sensors, Journal Year: 2024, Volume and Issue: 24(22), P. 7320 - 7320
Published: Nov. 16, 2024
IoT devices with limited resources, and in the absence of gateways, become vulnerable to various attacks, such as denial service (DoS) man-in-the-middle (MITM) attacks. Intrusion detection systems (IDS) are designed detect respond these threats environments. While machine learning-based IDS have typically been deployed at edge (gateways) or cloud, must be embedded within sensor nodes themselves. Available datasets mainly contain features extracted from network traffic (e.g., Raspberry Pi/computer) cloud servers. We developed a unique dataset, named Detection Smart Homes (IDSH) which is based on retrievable microcontroller-based devices. In this work, Tree-based into smart thermostat for real-time intrusion detection. The results demonstrated that achieved an accuracy 98.71% binary classification inference time 276 microseconds, 97.51% multi-classification 273 microseconds. Real-time testing showed capable detecting DoS MITM attacks without relying gateway cloud.
Language: Английский
Citations
1
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 78 - 86
Published: Jan. 25, 2024
The importance of Quality Service (QoS) remains utmost in the endeavor to provide high-quality network services. This study focuses on important area Specifically, it explores QoS-Aware Routing and Resource Allocation techniques, with a particular emphasis Class-Based Weighted Fair Queuing (CBWFQ). Our research utilizes NS-3 simulator thoroughly assess performance by analyzing crucial parameters such as latency, throughput, reliability. We draw insights from CAIDA Anonymized Internet Traces dataset. CBWFQ, an advanced queuing mechanism, is highlighted for its capability intelligently categorize prioritize traffic into separate classes, each customized weightings resource guarantees. outcomes derived our experimentation demonstrate significant enhancements reliability across various scenarios, confirming efficacy CBWFQ optimizing allocation guaranteeing superior QoS. not only tackles immediate difficulties encountered administrators, but also provides valuable service providers researchers aiming enhance face diverse patterns. In addition, we propose potential areas future investigation, including examination AI-driven QoS mechanisms adaptable strategies that can effectively navigate constantly changing environments. incorporation methodologies cutting-edge technologies, 5G iterations, presents promising opportunity improve management upcoming era.
Language: Английский
Citations
0
Deleted Journal, Journal Year: 2024, Volume and Issue: 19(2), P. 58 - 67
Published: Jan. 25, 2024
Modern smart cities face significant mobility difficulties, and the combination of Intelligent Transportation Systems (ITS) Vehicular Networks (VN) within context Internet Vehicles (IoV) promises a transformative approach to tackling these challenges. This abstract captures core this ground-breaking approach. Traffic congestion, environmental challenges, road safety are crucial considerations in cities. management systems automobiles can communicate real-time data thanks support provided by vehicular networks. By incorporating into larger IoT ecosystem, expands connection broadens range available services applications. study introduces novel Transport System designed for network traffic based on The machine learning models used build system Decision Tree (DT), Support Vector Machine (SVM), Neural Network, K-Nearest Neighbours (KNN), Naive Bayes. simulation results show system's effectiveness producing astonishing through thorough review. In particular, it maintains computing efficiency while achieving noteworthy level detection accuracy. success be due skilful use feature selection ensemble approaches, which together improve performance. summary, research provides state-of-the-art that makes enhance control IoV-based vehicle networks city scenarios. comparing different model intelligent CNN leads with 98.87% followed other methods as discuss result section. It also promising development field transportation because not only improves accuracy but ensures efficiency.
Language: Английский
Citations
0
Deleted Journal, Journal Year: 2024, Volume and Issue: 20(1s), P. 366 - 378
Published: March 28, 2024
The rapid expansion of Internet Things (IoT) devices has resulted in an unparalleled surge the production data and interconnectivity. Nevertheless, as IoT ecosystems become increasingly intricate, security concerns utmost importance, particularly access control systems. objective this research is to improve by utilizing a hybrid model for analyzing threats modeling attacks based on activities. This study two primary objectives: a) A classification used predict labels (attack or not) binary with impressive accuracy 98.18%. b) Another employed types M2M communication, achieving commendable 90%. goal create assess classification. will differentiate between regular system behavior malicious schemes (IoT). model, which combines strengths Gated Recurrent Units (GRU) Long Short-Term Memory (LSTM) networks, achieves exceptional rate model's high demonstrates its effectiveness precisely detecting potential minimizing false positives, thereby establishing strong basis improving security. second focuses complex area security, categorizing distinct forms Machine-to-Machine (M2M) communication within framework. employing both GRU LSTM remarkable accomplishment showcases aptitude distinguishing different attacks, including Distributed Denial Service (DDoS) Man-in-the-Middle attacks. provides professionals valuable insights proactively respond diverse accurately classifying attack types. strengthens overall posture Overall, offers thorough efficient combination threat analysis activity-based enhance IoT. obtained accuracies prediction highlight practical usability suggested systems against evolving cyber
Language: Английский
Citations
0