@kprcas.ac.in
Professor & Head,Department of Computer Science with Computer Science with Data Analytics
KPR College of Arts Science and Research
M.C.A.,M..,
Computer Science, Computer Science Applications, Information Systems, Software
Scopus Publications
M. P. Swapna and G. Satyavathy
Chapman and Hall/CRC
M. P. Swapna and G. Satyavathy
Springer Nature Singapore
Swapna M P and Satyavathy G
Seventh Sense Research Group Journals
Swapna M P and G. Satyavathy
EverScience Publications
Wireless Sensor Network (WSN) is a collection of low energy sensor nodes deployed in hostile complex environments. Their functionality gathers requisite data from the environment and transmits it to the base station for further processing. To enhance the performance of WSN, sensor nodes with different energy levels, capabilities and functionalities are deployed, leading to Heterogeneous WSN (HWSN). The initial energy, energy consumption rate, and residual energy differ for each node in a heterogeneous WSN. Many algorithms were proposed to accomplish an energy-efficient steady HWSN, but the performance level is not satisfactory. This paper presents a novel integrated approach, Energy-Aware Optimal Clustering & Securing Routing (EAOCSR). The algorithm amalgamated three techniques optimal clustering, reliable routing and secured transmission, considering energy retention and network lifetime as the vital parameters. Unequal clustering scheme, trust-based reliable and secure routing forms the core of EAOCSR. The performance of EAOCSR is analyzed using MATLAB simulations. It reveals that the proposed routing protocol EAOCSR has superior performance to existing protocols regarding energy utilization, throughput, network lifetime, stability and security. Index Terms – HWSN, Unequal Clustering, Trust, Blockchain, Stability, Security.
P. Dhanalakshmi and Dr.G. Satyavathy
Institute of Advanced Scientific Research
P. Dhanalakshmi and G. Satyavathy
American Scientific Publishers
The quality of images is decreased by noises. There exist several chances for the noises to occur while capturing and transmission the image. Noise removal becomes a thrust area of research in image processing. The outcome of the noise removal shows the quality of digital image processing techniques. Noises in image lead to the semantic gap problem in medical image processing. Semantic gap problem becomes a serious issue in the classification of the medical image. With the aim to overcome this issue, this research work proposes an efficient noise removal method based on relevant vector machine. Instead of using unsuited linear filters to detect noises, this research work uses the nonlinear filters which suit well to detect noises in multiple scale layers. The proposed method is applied to ADL dataset for the detection of lung cancer. The results clearly show that the proposed noise removal based relevant vector machine performs better in terms of accuracy.