Asian Journal of Convergence in Technology
Not a member yet
868 research outputs found
Sort by
Clustering Based Prediction for VM Workload in Green Computation
Get PDFIn order to support the different new generation equipment and technologies, cloud computing is depends to deal with the bulky data. Because a rich amount of data is generated using these devices and processing such big data need the cloud servers which can scale the computational ability according to demand. On the other hands to perform computation we need huge power supply and cooling system that increases the power consumption and emission of harmful gases. Thus, need to achieve green computing by reducing power consumption of computational cloud. In this context, we found VM(virtual machine) workload scheduling can be a good strategy to efficiently utilize the computational resources and reducing power consumption of cloud server.
Basically, the physical machines contain a number of virtual machines (VMs). These VMs are used to deal with the workload appeared for processing. If we better utilize the resources then we can process large number of jobs in less amount of VMs. Additionally, we can also turn off the ideal machines to reduce the power consumption. In this context the proposed work is motivated to work with VM scheduling techniques to achieve green computing. In recent literature we also identified that there are two kinds of VM scheduling approaches active and proactive. The proactive technique is more effective as compared to active approaches, due to prior knowledge of the workload on VM. So, in this paper we proposed green cloud predictive model for VMs workload using unsupervised learning (i.e clustering) like K-Mean, K-Medoid, Fuzzy C-Mean (FCM),Self-Organizing Map(SOM) to predict the future workload for VM’s scheduling and find the efficient clustering among them for workload prediction in view of green computing. The efficiency of clustering-based prediction is measured on parameters like accuracy, error rate
From Paper to Digital: Transitioning Employee Processes with an Employee Management System
Get PDFTeachers and students form the core of any educational institution, be it a college or school. However, traditional methods of managing their tasks often result in cumbersome paperwork and complex processes. This paper delves into the creation of a cost-effective system aimed at alleviating these challenges. Unlike many existing computerized systems that primarily focus on attendance, leave, and salary management, this proposed system seeks to address a broader scope. Typically, crucial employee data, including personal information, task assignments, leave records, and work allocations, are manually handled. To streamline these operations, a web-based Employee Management System is suggested herein. This system not only saves considerable time but also ensures accurate pay calculations, thereby enhancing efficiency. In contrast to the technologies discussed in existing literature, this solution offers a more user-friendly approach. The primary objective of this endeavor is to evaluate and enhance employee performance within the institution
Artificial Intelligence-Based Approaches for Condition Monitoring and Fault Diagnosis of Rolling Element Bearings”
Get PDFTo save money, improve dependability, and maintain system safety, rotating machinery must be regularly monitored. A variety of modern approach-based approaches are utilised to detect and predict faults in rolling element bearings. These techniques include data extraction, clever structures based on period and rate of recurrence, time-frequency domains and detail mix, sign/image processing, intelligent diagnostics, and statistics fusion. The prominence of AIML ideas has heightened interest in this subject. The application of artificial intelligence approaches to industrial equipment, mechanisation, and development represents the ultimate limit of AI adaptability. Signal and data processing techniques are employed to solve problems in a well-developed body of literature. This paper's main contribution is to provide a detailed review. Third, utilising emerging developments in artificial intelligence and techniques, fault detection methods employing time domain and frequency domain analysis, and the bearing's CM, which encompasses a variety of CM approaches
CRPA for Anti-jamming Capability
Get PDFControlled Reception Pattern Antennas (CRPAs) are designed to optimize the reception and processing of GPS/GNSS signals, minimizing interference while maximizing accuracy. By leveraging multiple antenna elements and advanced signal processing techniques, CRPAs achieve these objectives effectively. These systems are increasingly being adopted, particularly in the Defense sector. CRPAs are highly efficient in countering jamming and spoofing, as they dynamically adjust to mitigate such disruptive signals. To implement null-steering and beamforming techniques, it is essential to determine the direction of the interference or jammer. With the widespread use of GPS in defense equipment, ensuring reliable GPS performance in mission-critical operations, especially in jamming environments, is vital. As such, studying this technology within systems that can provide anti-jamming capabilities for GPS-enabled devices—such as drones, GPS receivers, target acquisition systems in firearms, tanks, and helicopters—is crucial
A DEEP AND MACHINE LEARNING COMPARATIVE APPROACH FOR NETWORKS INTRUSION DETECTION
Get PDFIntrusion detection is intergral section of firewalls and other attacks prevention applications that works side by side with the attack pouncing section. The strongest attack prevention application is that of wide range of attack pouncing capability. Recently, data driven models are used for this task which offers the required capability of multiple type of attack detection. In this paper, foucse given to establish an attack detection system that compatible with various datasets and able to draw similar perfromacne in attack flection. Multilayer perception (MLP), Convolutional neural network (CNN). Machine learning algorithms are also deployed such as Random Forest (RF) and Boosting algorithms such as XGBoost, AdaBoost and CatBoost. The MLP algorithm was realized with best intrusion detection performance, it yielded a higher accuracy in both dataset cases. Overall, the classification results on the UNSW-NB15 dataset suggest that machine learning algorithms can be successfully applied to network intrusion detection tasks, with various algorithms demonstrating high levels of accuracy in distinguishing between normal and malicious network traffic
Urban Traffic Detection for Autonomous Vehicles
Get PDFAutonomous vehicles leverage advanced sensors, artificial intelligence, and automation, enabling self-navigation without human intervention. These vehicles hold the potential to significantly improve road safety, enhance efficiency, and revolutionize transportation systems by reshaping how vehicles perceive, interpret, and respond to their environment.
The demand for such vehicles arises from the desire for improved urban planning, decreased parking needs, and flexible public transportation. Automation reduces errors, optimizes traffic flow, and produces favorable economic results.
This study underscores the crucial importance of advanced traffic and lane detection in reinforcing the reliability and safety of autonomous vehicles, playing a pivotal role in their ongoing evolution.
The proposed system operates in real-time, employing dynamic traffic data to inform decision-making. It integrates inputs from cameras, processing parameters such as lane positions, obstacles, and traffic symbols. A centralized control system, comprising Raspberry Pi and Arduino as master-slave components, employs specialized models for lane, object, and traffic symbol detection. This architecture guarantees continuous real-time decision-making and optimizes resource allocation, promoting a resilient and adaptive autonomous driving paradigm. The comprehensive nature of this approach not only aligns with contemporary transportation requirements but also proactively tackles the challenges anticipated in the future urban mobility landscape.
 
CONNECTING THE DOTS: LINKING CORONARY DISEASES WITH COVID-19 PATIENTS THROUGH SUPPORT VECTOR MACHINE ALGORITHM
Get PDFThe COVID-19 pandemic has left a lasting impact on global health, with a significant portion of survivors experiencing persistent health effects termed as ‘LONG COVID’ or ‘POST COVID-19 SYNDROME’. In this research, we propose a novel approach utilizing Support Vector Machine (SVM) algorithm to analyse patient data and predict the multifaceted nature of post-COVID-19 Syndrome, particularly focusing on the interlinkage of coronary diseases with COVID-19 patients. Our methodology involves collecting and analysing extensive patient data, including pre-conditions and post-conditions of COVID-19, to identify patterns and associations between various health issues. By leveraging the high-dimensional capabilities of SVM, we aim to provide accurate predictions and insights into the long-term health complications of COVID-19 survivors, thereby contributing to a better understanding of this critical area of healthcare. This approach stands out due to its ability to handle nonlinear relationships, noise in data, and large datasets effectively, offering valuable insights for healthcare professionals in managing post-COVID-19 complications
Al-Based Signal Intelligence for Real-Time Threat Detection
Get PDFTechnology in AI and signal processing has changed signal intelligence (SIGINT) in recent years. This study examines AI-based Signal Intelligence (AI-SIGINT) systems for real-time threat detection in military, cyber security, and critical infrastructure protection. AI-SIGINT uses cutting-edge machine learning (ML) and deep learning (DL) algorithms to evaluate massive volumes of signal data from radio frequency (RF), satellite, and mobile networks to detect and neutralize threats in real time. AI-SIGINT systems autonomously monitor, intercept, and decode signal communications to quickly identify aberrant patterns that may indicate hostile activity or impending threats. A key component of AI-based signal intelligence is adaptive danger detection. Using reinforcement learning (RL) and anomaly detection, the system continuously evolves to improve threat perception. This adaptability detects sophisticated, changing threats like jamming attempts, frequency hopping, and cyber intrusions. This research also examines AI-driven SIGINT's ethical issues, including data privacy and unlawful surveillance. It also addresses technology issues like merging AI algorithms with SIGINT infrastructure and the necessity for high computational resources
Impact of Consumer Price Index (CPI) and Inflation on GDP: Evidence from Bangladesh
Get PDFThe study examines the impact of the Consumer Price Index (CPI) and Inflation on the Gross Domestic Product (GDP) of Bangladesh during 2006-2007 to 2021-2022. The research investigates the connection between GDP and these economic indicators using multiple linear regression analysis. The results reveals that the CPI significantly boosts GDP, suggesting that shifts in consumer prices over time have a big impact on the nation's overall economic development. In contrast, there appears to be a modest negative correlation between the GDP and the inflation rate. The study emphasizes the CPI's crucial influence on Bangladesh's GDP and highlights its significance as a major factor influencing the country's economic growth. These findings have important ramifications for stakeholders and policymakers, emphasizing the necessity of taking calculated risks to control inflation and the CPI in order to promote Bangladesh's sustainable economic progress and enhance excellence of life for its populace
COMPARITIVE STUDY OF ALGORITHMS
Get PDFEffective sentiment analysis of multilingual social media data is crucial for grasping user sentiments across various linguistic contexts. This research explores the challenges and advancements in sentiment analysis techniques, particularly in environments with limited resources. While most studies focus on monolingual analyses, recent developments in deep learning, especially transformer models, have shown promise for multilingual applications. The study evaluates different frameworks for sentiment analysis, emphasizing essential steps such as data collection, preparation, feature extraction, and model selection to address linguistic diversity. It also examines various methods, including artificial intelligence techniques and multilingual approaches, assessing their effectiveness in low-resource settings. The goal is to validate the robustness of these frameworks and identify best practices for accurate sentiment analysis in constrained environments, ultimately enhancing global sentiment understanding through adaptable and advanced techniques.
Understanding user sentiments on social media is vital across diverse languages, especially in resource-limited situations. Although most research has centered on monolingual contexts, advancements in deep learning transformers have demonstrated effectiveness. Major social media platforms like Twitter and Facebook play a key role in extracting valuable insights from their vast and evolving data. This study investigates and evaluates cutting-edge sentiment analysis techniques, focusing on their performance in low-resource linguistic environments where data availability is limited. By conducting a comparative analysis of various models, the research seeks to confirm the robustness of these frameworks and identify the most effective techniques for sentiment analysis under linguistic constraints