Archive Browsing VOLUME 10 ISSUE 07 FEBRUARY 2022

Performance Evaluation of Hybrid Relay Selection in Cooperative Communication System
Abstract: Energy detectors, feature detectors, and cooperative sensing are just a few examples. During each sensing period, one or more secondary users (SUs) execute sensing on a single and the same channel in these schemes. This method of multiple SUs detecting a single channel at the same time may severely limit sensing efficiency. This paper presents a cooperative spectrum sensing system based on relays. MATLAB 2014a is used to simulate the suggested study. The impact of signal to noise ratio, chance of detection, and throughput on the suggested algorithm has been assessed.

Authors: Vaibhavi Sanjiv Pawar, Umesh Bhimrao Pagare

File Name: Vaibhavi_10070-22-101.pdf
 Download Abstract | Download Full Paper

Energy Efficient Clustering Algorithm in Wireless Sensor Networks using Genetic Algorithm
Abstract: Sensor nodes in a heterogeneous Wireless Sensor Network (WSN) have different capabilities, such as computational power and sensing range. When compared to homogeneous WSN, heterogeneous WSN layout and topology control are more confusing. Distinctive energy-efficient clustering methods for wireless sensor networks systems are considered, with clustering method, location awareness, heterogeneity level, and clustering features all being considered. However, each protocol is incompatible with heterogeneous WSNs. In this research, we put the Low-Energy Adaptive Clustering Hierarchy (LEACH) and the Genetic Algorithm (GA) optimized-LEACH to the test in a few different scenarios where high level heterogeneity is held to a minimum. To bring the conduct of these disparate protocols to a stop.

Authors: Pawan Panditrao Jagdale, Sarang Dagajirao Patil

File Name: Pawan+_10070-22-102.pdf
 Download Abstract | Download Full Paper

Credit Card Fraud Detection using BPSO based Features and Random Forest Classifier
Abstract: Credit card fraud is a social problem that faces many ethical challenges and poses a serious threat to businesses around the world. Machine learning algorithms are used to detect fraudulent transactions by authors. This study presents an implementation of an automated credit card fraud detection system where pre-processing the data, among others. Binary particle swarm optimization (BPSO) algorithms are used for the selection of features with a random forest (RF) classifier for training and testing from the Kaggle dataset. Sensitivity, precision, f-score, and accuracy are used as performance evaluation tools to evaluate the proposed technique.

Authors: Dr. Hemant N. Patel, Dr. Amit N. Patel, Mr. Sunil P. Patel

File Name: Hemant_10070-22-105.pdf
 Download Abstract | Download Full Paper

Heterogeneous Clustering using Modified Stable Election Protocol (M-SEP) in WSN
Abstract: Small networks and inexpensive communicating sensors are rapidly being employed in industrial applications and environmental monitoring, thanks to significant advances in technological development in recent years, particularly in microelectronics and wireless communication technology. However, the utilization of wireless sensor networks in such applications is constrained by sensor limitations such as processing capacity, memory size, and energy consumption or the networks own constraints, such as the limited capacity, dynamics caused by topological variation, and the proper communication protocols suited to this type of network. In this paper, we put the Modified-Stable Election Protocol (M-SEP) and the Low-Energy Adaptive Clustering Hierarchy (LEACH) to the test in a variety of scenarios where high-level heterogeneity is presented.

Authors: Vaishali Bhausaheb Amale, Rahul Manohar Patil

File Name: Vaishali_10070-22-103.pdf
 Download Abstract | Download Full Paper

Iris Recognition using Random Forest Classifier
Abstract: This study discusses the use of iris recognition for biometric identification in situations that demand a high level of security, such as prisons. The human iris database photographs are processed using an image processing approach. It determines the center coordinates as well as the radius of the iris. The iris image is also subjected to noise reduction. The collected features are used as inputs for the random forest classifier, which generates a class for the identification of the person. A hybrid approach to feature extraction is proposed in this study which uses a combination of Gabor wavelet and Harris Corner methods. The collected characteristics are classified using a random forest classifier. The simulation results are put to the test against publicly available data.

Authors: Prasad Satosh Padalkar, Lalit Prakash Patil

File Name: Prasad_10070-22-104.pdf
 Download Abstract | Download Full Paper

© 2016 IJDACR Journal. All rights reserved.