@acet.ac.in
PROFESSOR -ECE
ADITYA COLLEGE OF ENGINEERING AND TECHNOLOGY SURAMPALEM AP
Received Ph.d in CEG CAMPUS, ANNA UNIVERSITY, CHENNAI
3 DIMENSIONAL VLSI, EMBEDDED SYSTEMS - IOT, NANO - SOLID STATE DEVICE MODELS
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
V Balaji, T S Karthik, N Akiladevi, S. Sathya, V Mahalakshmi, and D Anup Kumar
IEEE
The quality of life of a person can be severely diminished by movement disorders such as Parkinson’s disease or essential tremor. Although deep brain stimulation (DBS) has emerged as a promising therapeutic strategy, there are still gaps in our ability to properly optimize therapy with the tools at our disposal. This study employs state-of-the-art NeuroAI technology to completely modify the way in which movement disorders are treated. The inability to make real-time adjustments to DBS settings in response to changes in the patient’s health is at the heart of the problems that plague current approaches. Traditional approaches typically employ fixed parameters that do not take into account individual differences in how they feel. This rigidity might cause unwanted consequences and subpar performance. NeuroAI, a complex AI system designed to interpret brain signals and individual patient data, lies at the heart of our approach. It permits continuous modifications to stimulation settings based on real-time study of patient reactions and symptom variations. Our method does this by continuously adapting to the patient’s changing requirements. Patients have reported dramatic improvements in symptom management, decreased side effects, and enhanced quality of life, as shown by the study’s promising early results. With the help of NeuroAI, DBS may be administered with unparalleled accuracy, giving patients new hope for a better, symptom-free future. This study is a major step forward in the direction of making deep brain stimulation a regular treatment for movement disorders that is both individualized and extremely effective.
T. S. Karthik, D. Kamalakkannan, S. Murugesan, Jyoti Prasad Patra, Md. Abul Ala Walid, Kalagotla Chenchireddy, Syed Musthafa A, and B. Jagadish Kumar
Informa UK Limited
C.N. Ravi, T.S. Karthik, K. Manikandan, Pcd Kalaivaani, Priyanka Nandkishor Chopkar, and Aviral Srivastava
IEEE
The tremendous growth of internet technology has drastically improved the large amount of connected devices. To secure network infrastructure from the damage that cyberattacks might cause, this has made an enormous attack surface that needs the deployment of practical and effective counter measures. In the contemporary era of active network transmission and throughput, Intrusion Detection System (IDS) plays a vital role in ensuring secure network resource and data from outside invasion. In recent times, IDS becomes an essential tool to enhance the efficiency and flexibility for unpredictable and unexpected invasions of the network. Deep learning (DL) is a well-known and essential method to resolve challenges and could learn rich features of massive information. Therefore, the study focuses on the design and development of the Cauchy Grasshopper Optimization Algorithm with Deep Learning for Cloud Enabled IDS (CGOA-DLCIDS) technique. The presented CGOA-DLCIDS method aims to recognize the presence of intrusions in the cloud platform. To achieve this, the CGOA-DLCIDS technique performs feature subset election by CGOA which reduces the feature subset and enhances the intrusion detection rate. Next, the CGOA-DLCIDS technique employs attention based long short-term memory (ALSTM) module for automated and accurate intrusion detection and classification. The simulations analysis of the CGOA-DLCIDS method on benchmark dataset highlighted the increasing results compared to recent IDS approaches.
Jaya Dipti Lal, T. Balachander, T.S. Karthik, Sandy Ariawan, Pratap M S, and Mohit Tiwari
IEEE
In recent times, the internet of Things (IoT) is an alternative model that is quickly getting ground in the scenario of current wireless telecommunication. Wireless sensor network (WSN) is a significant part of IoT, and it is primarily accountable for reporting and acquiring information. As coverage area and lifetime of WSN directly define the performance of IoT, how to design a technique for conserving node energy and decreasing node death rate becomes crucial problem. Sensor network clustering is an efficient technique to overcome this problem. It splits nodes into clusters and chooses one to be cluster head (CH). The data communication and transmission within single cluster are accomplished by its CH. This study develops a hybrid evolutionary algorithm-based energy efficient cluster head selection (HEA-EECHS) technique in the IoT environment. The presented HEA-EECHS technique concentrates on the effectual choice of CHs in the IoT environment. To do so, the HEA-EECHS technique derives an improved artificial jellyfish search algorithm (IAJSA) by the incorporation of oppositional based learning (OBL) approach into the traditional AJSA. Along with that, the HEA-EECHS technique designs a fitness function incorporating four parameters namely energy, cluster node density, average neighboring distance, and average distance to BS. The experimental assessment of the HEA-EECHS technique is investigated under several IoT nodes and the final results gives the value of 500 WMNs, the HEA-EECHS method has attained decreased CMO of 0.0015. The simulation output highlighted the improvised efficacy of the HEA-EECHS technique.
T. S. Karthik, Naziya Hussain, N K Anushkannan, Rajasekhar Pinnamaneni, Vijayakrishna Rapaka E, and Shyamali Das
IEEE
Intracranial haemorrhage (ICH) refers to a pathological disorder that requires quick decision-making and diagnosis. Computed tomography (CT) can be accurate and dependable diagnosis method for identifying haemorrhages. Automated recognition of ICH through CT scans with a computer-aided diagnosis (CAD) method will be useful to classify and detect the distinct grades of ICH. Due to the latest development of deep learning (DL) techniques in image processing applications, numerous medical imaging methods use it. Thus, this article develops an automated ICH detection and classification using Rider Optimization with Deep Learning (ICHDC-RODL) model. The presented ICHDC-RODL technique mainly determines the presence of ICH using DL concepts. In the presented ICHDCRODL technique, the features are generated by the use of Xtended Central Symmetric Local Binary Pattern (XCS-LBP) model. Moreover, the bidirectional long short-term memory (BiLSTM) method is employed for ICH diagnosis. At last, the rider optimization algorithm (ROA) is exploited for the hyperparameter tuning procedure of the BiLSTM method. To demonstrate the enhanced outcomes of the ICHDC-RODL technique, a series of simulations were performed and the results are examined under various aspects. The simulation outcomes indicate the enhancements of the ICHDC-RODL technique over recent approaches.
A. Praveena, N. Senthamilarasi, T. S. Karthik, Abirami S.K, Vijayakrishna Rapaka E, and Shyamali Das
IEEE
Autism Spectrum Disorder (ASD) is a developing disorder if the symptoms develop obvious in the initial years of age but it could be present in some age groups. ASD is mental health problem that affects communicational, social, and non-verbal performances. It could not be cured entirely but is decreased when identified initially. The primary analysis was hampered by the difference and severity of ASD symptoms and containing symptoms usually realized in other mental health problems as well. With the application of machine learning (ML) for the predictive and recognition of several diseases with optimum accuracy, a ray of hope to initial recognition of ASD dependent upon many physiological and physical parameters is projected. This article designs an Equilibrium Optimizer with Deep Learning Model for Autism Spectral Disorder Classification (EODL-ASDC) technique. The presented EODL-ASDC technique mainly focuses on the identification and classification of ASD. To attain this, the presented EODL-ASDC technique exploits the deep belief network (DBN) system to perform the classification procedure. In addition, the EO algorithm is employed for the optimal hyperparameter tuning of the DBN approach. To demonstrate the enhanced ASD classification result of the EODL-ASDC approach, an extensive range of experimental evaluates was executed. The experimental results demonstrate the improvements of the EODL-ASDC technique over other approaches.
T. S. Karthik, R.V. V. Krishna, T. K. Ramakrishna Rao, V. Manoranjithem, S. Kalaiarasi, and B. Jegajothi
IEEE
Content based image retrieval (CBIR) is commonly utilized in several application areas due to the rising significance of images in day to day lives. On comparing with textual data, images require high storage area and processing complexity. The latest advances in machine (ML) and deep learning (DL) models can be utilized for the design of effective CBIR In this view, this paper presents an evolutionary optimization algorithm on CBIR system using handcrafted features with squeeze networks (EOCBIR-HFSN) technique. The goal of the EOCBIR-HFSN technique is to proficiently retrieve the related images based on the query image (Q1). The proposed EOCBIR-HFSN technique involves the feature extraction process by the use of local binary patterns (LBP) based handcrafted features and SqueezeNet based deep features. Besides, the hyper-parameter tuning of the SqueezeNet model is performed by the grasshopper optimization algorithm (GOA), shows the novelty of the work. Finally, Euclidean distance metric is used to determine the highly similar images from the database. The comprehensive result analysis of the EOCBIR-HFSN technique take place on benchmark database reported the enhanced outcomes over the other techniques.
C. Pavithra, Pooja Singh, Venkatesa Prabhu Sundramurthy, T.S. Karthik, P.R. Karthikeyan, John T. Abraham, K.G.S. Venkatesan, Susheela Devi B. Devaru, and T.C. Manjunath
Elsevier BV
T. S. Karthik, K. Loganathan, A. N. Shankar, M. Jemimah Carmichael, Anand Mohan, Mohammed K. A. Kaabar, and Safak Kayikci
Hindawi Limited
This work addresses 3D bioconvective viscoelastic nanofluid flow across a heated Riga surface with nonlinear radiation, swimming microorganisms, and nanoparticles. The nanoparticles are tested with zero (passive) and nonzero (active) mass flux states along with the effect of thermophoresis and Brownian motion. The physical system is visualized via high linearity PDE systems and nondimensionalized to high linearity ordinary differential systems. The converted ordinary differential systems are solved with the aid of the homotopy analytic method (HAM). Several valuable and appropriate characteristics of related profiles are presented graphically and discussed in detail. Results of interest such as the modified Hartmann number, mixed convection parameter, bioconvection Rayleigh number, and Brownian motion parameter are discussed in terms of various profiles. The numerical coding is validated with earlier reports, and excellent agreement is observed. The microorganisms are utilized to improve the thermal conductivity of nanofluid, and this mechanism has more utilization in the oil refinery process.
Madhan Mohankumar, A. N. Shankar, T. S. Karthik, R. Saravanakumar, Hemakesavulu Oruganti, S. Venkatesa Prabhu, and N. Rakesh
Hindawi Limited
This study was conducted to assess and compare the crack-healing ability of conventional electrical sintered and microwave sintered Al2O3/x wt. % SiC (x = 5, 10, 15, and 20) structural ceramic composites. The crack-healing ability of both conventional electrical sintered and microwave sintered specimens was studied by introducing a crack of ∼100 µm length by Vickers’s indentation and conducting a heat treatment at 1200°C for dwell time of 1 h in air. The flexural or bending strength of sintered, cracked, and crack-healed specimens was determined by three-point bending test, and the phase variations by X-ray diffraction and SEM micrographs before and after crack-healing of both the sintering methods were studied and compared. The results show that almost all the specimens recovered their strength after crack-healing, but the strength of microwave sintered Al2O3/SiC structural ceramic composites has been shown to be better than that of conventional electrical sintered Al2O3/SiC structural ceramic composites. The microwave sintered crack-healed Al2O3/10 wt. % SiC specimen shows higher flexural strength of 794 MPa, which was 105% when compared with conventional electrical sintered Al2O3/10 wt. % SiC and crack-healed Al2O3/10 wt. % SiC specimen. It was found by X-ray diffractogram that before crack-healing, all the conventional electrical sintered samples have SiO2 phase which reduce the crack-healing ability and microwave sintered samples with 15 and 20 wt. % SiC show lesser SiO2 phase and 5 and 10 wt. % SiC samples have no SiO2 phase before crack-healing. However, after crack-healing treatment, all the samples have distinct SiO2 phase along with Al2O3 and SiC phases. Microwave sintered Al2O3/10 wt. % SiC specimen cracks were fully healed which was evident in SEM micrographs.
K. Loganathan, Nazek Alessa, Ngawang Namgyel, and T. S. Karthik
Hindawi Limited
This study explains the impression of MHD Maxwell fluid with the presence of thermal radiation on a heated surface. The heat and mass transmission analysis is carried out with the available of Cattaneo–Christov dual diffusion. The derived PDE equations are renovated into ODE equations with the use of similarity variables. HAM technique is implemented for finding the solution. The importance of physical parameters of fluid velocity, temperature, concentration, skin friction, and heat and mass transfer rates are illustrated in graphs. We found that the fluid velocity declines with the presence of the magnetic field parameter. On the contrary, the liquid temperature enhances by increasing the radiation parameter. In addition, the fluid velocity is low, and temperature and concentration are high in Maxwell fluid compared to the viscous liquid.
Banumathi S., Karthik T. S., Sasireka M., Kiran Ramaswamy, Vishnu J., Yuvan M. K., Kavin R. R., and Sathish Kumar S.
Hindawi Limited
Epoxy resin mixed with rice husk ash and quartz powder increases its dielectric strength. This paper presents the dielectric properties of the press board coated with this epoxy mixture. In this work, the press board, which is used in the transformer, is coated with three components: epoxy resin, rice husk ash, and quartz powder. The nanometer-sized quartz powder and rice husk ash are mixed in the particular ratio with the epoxy resin. The mixture of epoxy resin, quartz powder, and rice husk ash is coated on both sides of the press board. The dielectric constant, volume resistivity, and Tan Delta (dissipation factor) of the coated press board are compared with the noncoated press board. The results reveal that the coated board is having high dielectric constant and volume resistivity when compared to the noncoated board.
Nazek Alessa, B. Venkateswarlu, K. Loganathan, T.S. Karthik, P. Thirupathi Reddy, and G. Sujatha
Hindawi Limited
The focus of this article is the introduction of a new subclass of analytic functions involving q-analogue of the Bessel function and obtained coefficient inequities, growth and distortion properties, radii of close-to-convexity, and starlikeness, as well as convex linear combination. Furthermore, we discussed partial sums, convolution, and neighborhood properties for this defined class.
Nazek Alessa, K. Tamilvanan, K. Loganathan, T. S. Karthik, and John Michael Rassias
Hindawi Limited
In this work, we examine the generalized Hyers-Ulam orthogonal stability of the quartic functional equation in quasi- β -normed spaces. Moreover, we prove that this functional equation is not stable in a special condition by a counterexample.
P. R. Karthikeyan, P. Sakthivel, and T. S. Karthik
Springer Science and Business Media LLC
Foreground detection plays a vital role in finding the moving objects of a scene. For the last two decades, many methods were introduced to tackle the issue of illumination variation in foreground detection. In this article, we proposed a method to segment moving objects under abrupt illumination change and analyzed the merits and demerits of the proposed method with seven other algorithms commonly used for illumination-invariant foreground detection. The proposed method calculates the entropy of the video scene to determine the level of illumination change occurred and select the update model based on the difference in entropy values. Benchmark datasets possessing different challenging illumination conditions are used to analyze the efficiency of the foreground detection algorithms. Experimental studies demonstrate the performance of the proposed algorithm with several algorithms under various illumination conditions and its low time complexity.
This paper presents the systemic approach to education and educational innovation practices. A new approach and alternate view of game model used to customize learning experiences. It uses intelligent data, analysis models to find how the students learn and improve on their experience. The way of teaching aims to deploy the animations and easy way of understanding the concepts through game based learning. The main aim is to design interactive tool usage for simplicity and less complexity. The proposed gaming strategy is deployed for the purpose of easy way of understanding to arrange in ordered fashion called Sequencer Square Box technique. Using this technique, players or learners can have the ability to find, share the information through friends. They can easily explore, collaborate their ideas and progress towards multiple literacies in any domains and readiness to build any challenges in future.
B. Prakash, S. Jayashri, and T. S. Karthik
Springer Science and Business Media LLC
Wireless mesh networks are a special class of wireless networks that are implemented as a collection of radio nodes in a mesh pattern or topology. Unlike MANETs, the mobility of nodes is very less in the topology. Quality of service is an essential metric in the performance of mesh networks which are attributed to several parameters including optimal routing through shortest path, ability for other nodes to communicate even if a particular node in the mesh fails, minimization of packet loss and time delay, computational complexity and cost, energy. This research paper is focused toward minimization of energy taken as the objective function and a five-stage neural network is used and trained after optimizing with a genetic algorithm. The experiments have been conducted in NS2 and Qualnet environment with a varying number of mesh routers and energy computed. The performance of energy savings has been compared against conventional routing techniques like AODV and a bee colony optimization technique presented in the literature. An energy savings of 51% have been reported in the paper justifying the superiority of the hybrid G-ANN algorithm.
Denny Mathew and T S Karthik
IEEE
The key factor making Delta-Sigma modulators (DSM) one of the most popular components in modern electronic circuits is its high linearity. This is achieved by using a high oversampling ratio which is unfortunately the limiting factor towards its application in high frequency circuits. The necessity of high processing speed and power, the increased cost and complexity and wastage of available bandwidth are some of the significant demerits of using a high oversampling ratio. This paper suggests that the delta sigma modulators require a high frequency processing and not high oversampling ratio. A parallel structure to perform the high frequency processing along with an adaptive method to improve the signal quality at the output is proposed. The suggested technique allows the simultaneous execution of fast and complex computations required for wireless systems. The analysis is performed using MATLAB simulations and the results claim a reduction in oversampling ratio by a factor of 16 while keeping the same signal to noise ratio. The proposed architecture is implemented on a field-programmable gate array (FPGA) board which is then validated with a code division multiple access signal. The output signal bandwidth is observed to be increasing four times without any increase in the sampling frequency.