SNSivakumar Nadarajan

Sivakumar Nadarajan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Sivakumar Nadarajan has a total of 21 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Tokenizing Power

    by Sivakumar Nadarajan, Mavlanova Barchinoy Shonazar Qizi, R. N. Ravikumar, Pardaev Jamshid, S. Srinivasan, Deemed To Be, S. Aarthi

    The tokenization of energy resources through blockchain technology represents a transformative step toward sustainable, transparent, and decentralized energy systems. By enabling fractional ownership, peer-to-peer energy trading, and democratized investment structures, tokenized models foster inclusivity and enhance capital mobilization in renewable energy markets. Case studies such as Power Ledger in Australia and WePower in Europe highlight tangible benefits, including reduced transaction costs, improved trading efficiency, and broader investor participation. Despite its promise, tokenized energy ecosystems face regulatory uncertainty, ethical dilemmas, and implementation challenges. Addressing these barriers through harmonized legal frameworks, technological innovation, and global governance can accelerate adoption while ensuring equity and energy justice. This study consolidates ownership models, performance metrics, and governance implications to provide a comprehensive roadmap for advancing tokenized energy systems.

  • Transfer Learning for IoT Applications

    by R. N. Ravikumar, S. Aarthi, Anand Singh, N. Sivakumar, S. Kavitha Bharathi, S. Sujitha

    The integration of IoT and AI enhances autonomous and adaptive systems, leveraging Transfer Learning and Explainable AI (XAI) for improved efficiency. This research explores AI-driven IoT applications, bridging theory and practice. Transfer Learning minimizes data needs and computational costs, increasing flexibility across smart cities, precision agriculture, healthcare, and energy management. AI models act as knowledge transfer nodes, enabling adaptive decision-making. XAI ensures transparency, interpretability, and trust in AI-assisted IoT, crucial for ethical compliance and reliable decision-making. Challenges include data processing, scalability, security, and privacy. The study addresses bias detection, fair data practices, and ethical AI deployment. Using interdisciplinary research and practical examples, it guides researchers, engineers, and industry professionals in advancing intelligent IoT systems.


  • Big Data and Cloud Computing for Climate-Oriented Remote Sensing

    by R. N. Ravikumar, S. Aarthi, Sivakumar Nadarajan, Gularam Masharipova, Zokir Mamadiyarov, S. Tamilselvi

    Improved climate change surveillance arises from rapid remote sensing via satellites, UAVs, and sensor networks. However, real-time data processing faces storage and format challenges due to growing data volumes. Deep learning enhances remote sensing for climate monitoring when integrated with big data tools in cloud environments. Scientists use multispectral, hyperspectral, and radar systems for analyzing surface temperatures, deforestation, sea levels, and glaciers. Research on Google Earth Engine's structure explores tool performance between AWS and Azure for geolocation-based deep learning. Image analysis benefits from cloud clusters and statistical models to detect climate anomalies. Cloud-powered systems monitor air pollution and floods. To ensure effectiveness, large-scale systems must follow protocols for data processing, optimization, and quality control. The section concludes with data privacy, then shifts to hybrid edge-cloud models and affordable real-time climate analytics. These methods inform policy with critical adaptation data.


  • Cryptography Securing Data in Motion Within Blockchain, Internet of Everything, and Federated Learning

    by S. Aarthi, K. Aravinthan, R. N. Ravikumar, N. Sivakumar, Soram Wanglen

    Data protection relies on cryptography to secure data across Blockchain, IoE, and Federated Learning systems. Strong cryptographic methods ensure confidentiality, authenticity, and integrity, safeguarding evolving digital security needs. Key techniques include symmetric and asymmetric encryption, hash functions, digital signatures, and zero-knowledge proofs. Cryptography enables secure protocols like TLS, homomorphic encryption, and differential privacy while addressing quantum-resistant security challenges, ensuring robust digital privacy solutions.


  • Top Cited Papers
    Open Access

    A Hybrid Brain Tumor Classification Using FL With FedAvg and FedProx for Privacy and Robustness Across Heterogeneous Data Sources

    by N. Sivakumar, Ahmad Raza Khan, Syed Umar, R. N. Ravikumar, I. Bremnavas, Munindra Lunagaria, Krunal Vaghela, Ghanshyam G. Tejani, Sunil Kumar Sharma

    Data privacy and heterogeneity among healthcare settings present fundamental challenges to machine learning (ML) brain tumor classification (BTC) model development based on local data. In this paper, we outline the need to develop an effective brain tumor diagnosis model while ensuring data security. We propose a distributed learning framework based on Federated Learning (FL) to build such a system. This framework combines Federated Averaging (FedAvg) and Federated Proximal (FedProx) to train Convolutional Neural Networks (CNN) on data hosted by multiple clients. FedAvg saves these locally trained models from the patients to make a global model while keeping sensitive patient data decentralized and private. Nevertheless, traditional FL suffers from model instability and divergence due to the disparity of data distributions among clients. FedProx addresses this challenge by introducing a proximal regularization term that penalizes divergence from the global model to stabilize local training considering data heterogeneity. In this work, we utilized the Brain Tumor MRI Dataset from Kaggle and trained the model to classify four brain tumors (BT), namely glioma, meningioma, pituitary tumor, and no-tumor class. Experiments were conducted for ten communication rounds, with each client training the model independently based on its local dataset and without sharing data. Results revealed that the method consistently outperformed accuracy, recall and F1-score, further validating the capabilities of the proposed framework in effectively handling data heterogeneity while conserving data privacy. The global model outperformed models that had been centralized and trained on a smaller, common dataset observed in FL scenarios to improve the potential for collaborative medical diagnosis. This method is especially suitable for medical applications with strict privacy regulations, as it provides an approach to linking heterogeneous data from multiple healthcare institutions.