International Journal on Recent and Innovation Trends in Computing and Communication
Not a member yet
8613 research outputs found
Sort by
Outlier Detection Mechanism for Ensuring Availability in Wireless Mobile Networks Anomaly Detection
Get PDFFinding things that are significantly different from, incomparable with, and inconsistent with the majority of data in many domains is the focus of the important research problem of anomaly detection. A noteworthy research problem has recently been illuminated by the explosion of data that has been gathered. This offers brand-new opportunities as well as difficulties for anomaly detection research. The analysis and monitoring of data connected to network traffic, weblogs, medical domains, financial transactions, transportation domains, and many more are just a few of the areas in which anomaly detection is useful. An important part of assessing the effectiveness of mobile ad hoc networks (MANET) is anomaly detection. Due to difficulties in the associated protocols, MANET has become a popular study topic in recent years. No matter where they are geographically located, users can connect to a dynamic infrastructure using MANETs. Small, powerful, and affordable devices enable MANETs to self-organize and expand quickly. By an outlier detection approach, the proposed work provides cryptographic property and availability for an RFID-WSN integrated network with node counts ranging from 500 to 5000. The detection ratio and anomaly scores are used to measure the system's resistance to outliers. The suggested method uses anomaly scores to identify outliers and provide defence against DoS attacks. The suggested method uses anomaly scores to identify outliers and provide protection from DoS attacks. The proposed method has been shown to detect intruders in a matter of milliseconds without interfering with authorised users' privileges. Throughput is improved by at least 6.8% using the suggested protocol, while Packet Delivery Ratio (PDR) is improved by at least 9.2% and by as much as 21.5%
Deep Learning-Based Big Data Analytics Model Based on Teaching Reforms in Three-Dimensional Composition
Get PDFWith the development of online education and big data analysis, new teaching models and methods have emerged. The integration of online and offline teaching modes based on big data analysis has become an effective way to promote teaching reform and practice in the field of three-dimensional composition. It is important to incorporate teaching reform into the teaching of three-dimensional composition to improve the quality of education and better prepare students for their future careers. This paper evaluated the contribution of teaching reform to the improvement of student performance. This paper designed a Deep Learning (DL) big data analytics model for data clustering and classification. The student performance is monitored for both online teaching and offline teaching classes. The collected data is clustered with the directional clustering process for the computation of feature space. With the estimated feature space value Hidden Markov Model (HMM) is implemented for the estimation of statistical data derived from the feature spaces. The extracted data were applied over the RESENT- 50 model for the classification of students’ performance. The data analysis with DL model stated that student performance in offline teaching is more significant than offline teaching in 3-dimensional aspects
Empowering Visually Impaired through the Assistance of SAHAYAK – A Walking Aid for the Blind
Get PDFTo help blind people overcoming difficulty in their movement in the physical environment and even in their home, a study on an engineering concept is very much necessary. So, our research comes out with an aid that will help blind people in their surroundings. It can detect any obstacle that will block the path of the blind. And The motion of the user can be sensed by the bot. Thus, Blind people can comfortably receive the help of our bot in assisting their movement from one place to another. This paper describes about an automated vehicle which can be controlled by an ultrasonic sensor to avoid obstacles when they move in their environment. Our automated robotic system is made up of an ultrasonic sensor and Arduino micro controller controls our automated bot. It is located in the front part of the bot. The ultrasonic sensor retrieves the data from the environment through the sensors attached to the bot. When any obstacle is detected then immediately that path is changed and an obstacle free path is chosen. The bot wheel is moved based on the data received by the controller from the sensor. The direction and wheel movement of the bot and will be decided from the ultrasonic sensor sensing and also using wheel encoder. It is used for detection and avoidance of interference. The controller is also programmed to be used with an android application
Some Clustering Methods, Algorithms and their Applications
Get PDFClustering is a type of unsupervised learning [15]. When no target values are known, or "supervisors," in an unsupervised learning task, the purpose is to produce training data from the inputs themselves. Data mining and machine learning would be useless without clustering. If you utilize it to categorize your datasets according to their similarities, you'll be able to predict user behavior more accurately. The purpose of this research is to compare and contrast three widely-used data-clustering methods. Clustering techniques include partitioning, hierarchy, density, grid, and fuzzy clustering. Machine learning, data mining, pattern recognition, image analysis, and bioinformatics are just a few of the many fields where clustering is utilized as an analytical technique. In addition to defining the various algorithms, specialized forms of cluster analysis, linking methods, and please offer a review of the clustering techniques used in the big data setting
A Trust-Based Group Key Management Protocol for Non-Networks
Get PDFIn this paper, a secure and trust-based group key management protocol (GKMP) is presented for non-networks such as MANET/VANET. The scheme provides secure communication for group members in a dynamic network environment and does not restrict the users (registered or non-registered), allowing for flexible group communication. The proposed scheme is designed to address the challenges of key distribution, secure grouping, and secure communication. For result evaluation, first of all formal and informal security analysis was done and then compared with existing protocols. The proposed trust-based GKMP protocol satisfies the authentication, confidentiality of messages, forward/backward security concurrently as well as shows robustness in terms of packet delivery ratio and throughput
Optimized Visual Internet of Things in Video Processing for Video Streaming
Get PDFThe global expansion of the Visual Internet of Things (VIoT) has enabled various new applications during the last decade through the interconnection of a wide range of devices and sensors.Frame freezing and buffering are the major artefacts in broad area of multimedia networking applications occurring due to significant packet loss and network congestion. Numerous studies have been carried out in order to understand the impact of packet loss on QoE for a wide range of applications. This paper improves the video streaming quality by using the proposed framework Lossy Video Transmission (LVT) for simulating the effect of network congestion on the performance of encrypted static images sent over wireless sensor networks.The simulations are intended for analysing video quality and determining packet drop resilience during video conversations.The assessment of emerging trends in quality measurement, including picture preference, visual attention, and audio visual quality is checked. To appropriately quantify the video quality loss caused by the encoding system, various encoders compress video sequences at various data rates.Simulation results for different QoE metrics with respect to user developed videos have been demonstrated which outperforms the existing metrics
Blockchain-Based Data Storage for Secure Electronic Health Records: A Comprehensive Review
Get PDFThe integration of blockchain technology into Electronic Health Records (EHRs) represents a transformative advancement in the healthcare industry, offering solutions to longstanding issues of data security, privacy, and interoperability. Traditional EHR systems are plagued by vulnerabilities, including data breaches, unauthorized access, and inefficient data sharing. Blockchain, with its decentralized, immutable, and transparent architecture, provides a robust framework for secure health data management. This review paper explores the potential of blockchain technology in enhancing the security and efficiency of EHR systems. It discusses various blockchain-based frameworks and their applications in healthcare, highlighting how blockchain can ensure data integrity, patient control, and interoperability. The paper also examines the role of smart contracts in automating access control and data validation, thereby enhancing the privacy and security of patient information. Through a comprehensive analysis of recent literature, this review identifies the benefits and challenges of implementing blockchain in EHR systems and outlines future research directions to address emerging issues and improve system efficienc
The Role of Artificial Intelligence, Machine Learning, and Deep Neural Networks in Medical Imaging: Applications, Strengths, and Challenges
Get PDFThe integration of Artificial Intelligence (AI) in medical imaging has revolutionized the field of medical diagnostics, offering unprecedented accuracy and efficiency in disease detection and management. This review paper explores the current role and future potential of AI in medical image diagnosis, summarizing key findings from recent literature. AI techniques, particularly machine learning (ML) and deep learning (DL), have demonstrated remarkable capabilities in analyzing complex medical images, facilitating early detection of diseases, and aiding in clinical decision-making. The reviewed studies highlight AI's success in various medical domains, including oncology, neurology, cardiology, and radiology, where AI has enhanced diagnostic precision and personalized treatment planning. Despite these advancements, challenges such as data heterogeneity, the need for extensive validation, and ethical considerations persist, necessitating further research. This paper underscores the transformative impact of AI in medical imaging and calls for ongoing efforts to overcome existing barriers to fully realize its potential in clinical practice
Optimal Criteria Approach for Tracking Motion Detection in Video Surveillance Using Image Processing Techniques
Get PDFSecurity system like video surveillance or Close Circuit TeleVision installation in the premises is an essential component for monitoring, recollecting, and observing the events in its actual order for future reference and evidence to the proprietor hypothesis. The process of identifying and tracking motion detection in video surveillance CCTV or any video source system is a tideous process due to its dynamic frame modifications in the image analysis approach. The existing surveillance video motion detection approach methods lacks in the areas of hardware feasibility shortcomings, frame difference detection errors, back ground image analysis failures, unable to perform effective comparison and verification analysis. The main issues of false alarms, ghost , incorrect and irrelevant data plays its substantial role in degrading the performance of motion detection in video surveillance system. This research article proposes a image processing approach for handling motipn detection in video surveillance system using image processing techniqes. In near future this research article focuses on the implementation of soft computing based motion detection in video surveillance with augmented reality system
Preserving the Support of Sensitive Item(s) while Hiding Sensitive Association Rules
Get PDFAn essential data-mining method for identifying intriguing relationships among a sizable collection of data objects is association rule mining. It may be a threat to the privacy of uncovered confidential information since it may reveal patterns and other types of sensitive knowledge that are hard to obtain in other ways. Such data needs to be shielded against unwanted access. Numerous tactics had been put forth to conceal the knowledge. Some employ data disruption, clustering, data distortion, and distributed databases across multiple sites. The need to strike a balance between the user's legitimate needs and the secrecy of exposed data is a challenge with hiding sensitive rules that has not yet received enough attention. The suggested method makes advantage of the data distortion methodology, which modifies the sensitive elements' position without changing their support. The database is still the same size. It first prunes the rules using the concept of representative rules, and then it conceals the rules that are sensitive. This strategy has the advantage of hiding the maximum number of rules; in contrast, the existing ways are unable to conceal all the needed rules, which should be concealed in the fewest number of passes. The suggested method is also contrasted with current approaches in the paper