Bøker av Gunavathi Ramasamy

The rapid growth of wireless networks in advanced networking enhances comfort and security. Ad-hoc wireless networks, including Mobile Ad-hoc Networks (MANET), Vehicular Ad-hoc Networks (VANET). VANET, a subset of MANET, facilitates Vehicle-to-Vehicle (V2V) and Vehicle-to-Roadside (V2R) communications. Its components include On-Board Unit (OBU), Road Side Unit (RSU), and Trusted Authority (TA), with applications in Intelligent Transport System (ITS). VANETs face broadcast storms causing collisions and data redundancy. The proposed research focuses on reviewing existing broadcast algorithms, suggesting a cluster-based protocol to minimize storms, avoid redundant data, reduce transmission delays, and form enhanced clusters. Efficient cluster head selection is crucial. Protocols like Cluster-Based Efficient Broadcasting (CBE-B), Selective Reliable Communication (SRC), and Improved Cluster Head Selection (ICHS) address these issues, outperforming existing solutions.

The fundamental notions behind diabetes mellitus and its types, causes, diagnosis, and significance in the prediction of DM. This technical work gives a short introduction to Data Mining and Machine Learning techniques for the prediction of diabetes mellitus. The different methodologies, techniques, and methods involved in predicting diabetes disease. Additionally, current advances in machine learning were highlighted, which have had a substantial influence on the identification and treatment of diabetes. The entire automatic disease prediction system utilizes data collection, noise removal, feature extraction, selection, and classification techniques. The described steps are examined in different authors perceptively to get the better knowledge about particular steps. The class imbalance problem on medical data and applied sampling technique is evaluated with different iteration which denotes the number of sampling process. Adaptive sampling technique is adopted for balancing the class in medical dataset. This improves the performance of the gradient boosting classifier that provides 97.78% accuracy.

Ad hoc network is a set up anywhere without any need for external infrastructure such as wires of base station. However, there is an increasing need for connectivity in situations where there is no base station available there emerges ad hoc networks. They are often mobile and termed as MANET and it is an autonomous system of mobile nodes connected by wireless links. Each node operates as an end system and a router for all other nodes in the network. A good number work of work has been devoted over the past few years in designing scalable routing protocols for MANETs.

Network intrusion detection is one of the central systems used in cyber security to prevent the intrusions in the organisation's networks. Tackling the attempts to compromise the confidentiality, integrity and availability of computer networks' security mechanisms in a big data environment is the most challenging task due to the volume and variety of big data. This study presented to tackle the challenges in network Intrusion Detection Systems (IDS) and demonstrate intelligent algorithms' development to detect the intrusions in big network data. The problem is the practical selection of the features from the network dataset as it dramatically impacts the intrusion detection accuracy. Hence, an efficient feature selection approach must be introduced to achieve higher accuracy with a reduced number of features.