RPR. S. Ponmagal

R. S. Ponmagal

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


R. S. Ponmagal has a total of 55 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Explainable Water Quality Index Forecasting Framework Using Hybrid Graph Attention and Transformer Networks for Intelligent Early Warning Systems

    by R. Sarala, R.S. Ponmagal

    The importance of water quality monitoring for aquatic ecosystem protection cannot be overemphasized, but the traditional machine learning algorithms used for monitoring data analysis have shortcomings in failing to account for the spatiotemporal interactions of the data. This study seeks to develop a spatiotemporal deep learning framework that employs Graph Attention Networks (GAT), Temporal Fusion Transformers (TFT), and SHAP-based interpretability for WQI forecasting and early warning purposes. Firstly, preprocessing of multisite physicochemical data involves filling in missing values using interpolation, removal of anomalies, and Min-Max normalization. Secondly, a spatial graph of monitoring stations is created, whereby the GAT layer captures the spatial information by allocating attention scores to the neighboring nodes, and TFT is employed to capture long-term temporal interactions and nonlinear trends. Lastly, a persistence-based early warning system identifies pollution degradation risks, while SHAP gives insight into parameter contribution. Empirical tests have revealed that the GAT-TFT model significantly outperforms conventional techniques (ANN, LSTM, Hybrid LSTM, and basic GNNs), exhibiting superior prediction efficiency with MAE = 0.048, RMSE = 0.079, and 𝑅2 = 0.96. Moreover, the early warning system exhibits an accuracy of 0.92, precision of 0.89, recall of 0.94, an F1 score of 0.91, and a False Alarm Rate of 0.08, where dissolved oxygen, turbidity, and pH were found to be the most significant factors. In conclusion, the presented framework offers environmental agencies a powerful and highly efficient method for proactive management of water resources.

  • Open Access

    IDMN-SP: Improved Deep Maxout Network with Adaptive Structured Pruning for Resource-Efficient Federated Learning in IoT Big Data Analytics

    by Ch. Ellaji, R.S. Ponmaga, V. Saritha

    The exponential growth of Internet of Things (IoT) ecosystems generates massive heterogeneous big data, requiring efficient, privacy-preserving analytics that operate under strict edge resource constraints. Current federated learning (FL) frameworks often struggle with high computational costs, poor bandwidth, and ineffective pruning or ReLU activation bottlenecks that trigger severe accuracy losses and dead neurons. To address these challenges, this paper introduces the Improved Deep Maxout Network with Adaptive Structured Pruning (IDMN-SP), a lightweight federated learning architecture designed for resource-efficient IoT big data analytics. IDMN-SP features two core innovations: replacing traditional ReLU with a Bendable Linear Unit (BLU) function to eliminate dead neurons and enhance feature representation, and an adaptive gradient-based structured pruning mechanism employing normalized L2-norm thresholding to selectively remove redundant convolutional filters without discarding key features, paired with a targeted fine-tuning recovery phase. Evaluated on the CIFAR-10 benchmark dataset under non-IID conditions (𝛽 = 0.5), the proposed framework achieves an outstanding classification accuracy of 84.73%, a Precision of 84.41%, a Recall of 84.59%, and an F1-Score of 84.50%. Moreover, IDMN-SP drastically improves resource efficiency by reducing model complexity to 24.19 million FLOPs, shrinking model size to 1.94 MB, and lowering communication bandwidth to 1.94 MB per round, thereby outperforming baseline methods like FedDA by 4.56 percentage points in accuracy. These results demonstrate that IDMN-SP successfully bridges the gap between high predictive accuracy and strict resource conservation, making it an optimal solution for edge-based IoT environments.

  • Open Access

    Prediction of Water Quality Index (WQI) Using a Novel Hybrid Water Wave-Optimized Extreme Gradient Boostin

    by R Sarala, R.S Ponmagal

    Determining the drastic changes in water quality in this contaminated period is an urgent requirement that is crucial for the continued existence of the current and growing water demand. This is why it is essential for maintaining the water supply. Due to variability and complexity in the data, the water quality index (WQI) has been developed as a reliable way to express water quality. Analyzing the data relative to the water quality has been extremely difficult. In this investigation, the WQI prediction of water samples using the novel Hybrid Water Wave Optimized eXtreme Gradient Boosting (HWW-XGB). In this research, the data preprocessing using the Outlier Detection (Box Plot Analysis) and Min-max normalization was finished to minimize the computational time. Pre-processing is a crucial first step in the proposed system because the dataset contains inappropriate and inaccurate data. The Independent Linear Analysis (ICA) is utilized to extract the required features. The pre-processed features are removed through feature extraction, eliminating unnecessary data for the next essential phase. The collected features are fed into Hybrid Water Wave Optimized eXtreme Gradient Boosting (HWW-XGB) classifier. The effectiveness of the proposed approach was demonstrated when compared to the existing method.

  • Open Access

    An Intelligent Intrusion Detection System for VANETs Using Adaptive Fusion Models

    by M. Shanthalakshmi, R.S. Ponmagal

    Vehicular Ad Hoc Networks (VANETs) play a vital role in the development of Cyber-Physical Systems (CPS) to enable real-time communication for improving road safety and traffic efficiency. Due to the VANETs' decentralized and dynamic nature, they are prone to various types of cyber-attacks, including intrusion, spoofing, and denial-of-service (DoS) attacks. This article presents an Adaptive Fusion Intrusion Detection Model (AFIDM), a multi-level framework that uses machine learning techniques, such as Random Forest, XGBoost, Decision Trees, and K-Nearest Neighbor (KNN), to deal with such vulnerabilities. AFIDM also employs a dynamic weight adjusting mechanism and an adaptive feedback loop to adapt to the evolving threats and achieve better detection accuracy. AFIDM achieved 98.7% accuracy, 96.5% precision, and recall of 95.8% on the VeReMi dataset used for training and validation and outperformed other baseline models. With low latency and scalability, the proposed model presents a robust solution for real-time intrusion detection in VANETs for the secure and efficient operation of intelligent transportation systems


  • Open Access

    Optimizing Secure Communication in Intelligent Transportation Networks with Certificate-less Authorization in VANETs

    by M. Shanthalakshmi, R.S. Ponmagal

    As an outcome of technology progress, among the most widely used design models in intelligent transportation networks is Vehicular Ad-hoc Networks (VANETs). Assistance for Vehicle to Vehicle (V2V) connectivity is offered via VANETs. Automobile access to actual time traffic, weather, and transportation data is made possible via VANETs, which also assure traffic safety and effectiveness. The legitimacy of an organisation providing data is a crucial factor that must be taken seriously in order to ensure secure interaction in VANETs. In regards to safety mutual identification along with safe distribution of keys in VANETs continue to pose certain disadvantages considering that numerous protocols that have been suggested. However, since they depend on the authorised authority, they are susceptible to assaults through the firewall and Reliable Control (RC), which may raise safety issues amongst customers. we provide a certificate less secure authorization system that lessens the load on automobiles' internal storage and satisfies the delay criterion by doing away with certificate. The suggested system is strong and resistant against attacks from malevolent users


  • Retracted
    Open Access

    RETRACTED ARTICLE: An intelligent dynamic cyber physical system threat detection system for ensuring secured communication in 6G autonomous vehicle networks

    by Shanthalakshmi M, Ponmagal R S

    Smart cities have developed advanced technology that improves people’s lives. A collaboration of smart cities with autonomous vehicles shows the development towards a more advanced future. Cyber-physical system (CPS) are used blend the cyber and physical world, combined with electronic and mechanical systems, Autonomous vehicles (AVs) provide an ideal model of CPS. The integration of 6G technology with Autonomous Vehicles (AVs) marks a significant advancement in Intelligent Transportation Systems (ITS), offering enhanced self-sufficiency, intelligence, and effectiveness. Autonomous vehicles rely on a complex network of sensors, cameras, and software to operate. A cyber-attack could interfere with these systems, leading to accidents, injuries, or fatalities. Autonomous vehicles are often connected to broader transportation networks and infrastructure. A successful cyber-attack could disrupt not only individual vehicles but also public transportation systems, causing widespread chaos and economic damage. Autonomous vehicles communicate with other vehicles (V2V) and infrastructure (V2I) for safe and efficient operation. If these communication channels are compromised, it could lead to collisions, traffic jams, or other dangerous situations. So we present a novel approach to mitigating these security risks by leveraging pre-trained Convolutional Neural Network (CNN) models for dynamic cyber-attack detection within the cyber-physical systems (CPS) framework of AVs. The proposed Intelligent Intrusion Detection System (IIDS) employs a combination of advanced learning techniques, including Data Fusion, One-Class Support Vector Machine, Random Forest, and k-Nearest Neighbor, to improve detection accuracy. The study demonstrates that the EfficientNet model achieves superior performance with an accuracy of up to 99.97%, highlighting its potential to significantly enhance the security of AV networks. This research contributes to the development of intelligent cyber-security models that align with 6G standards, ultimately supporting the safe and efficient integration of AVs into smart cities.


  • Retraction
    Open Access

    Retraction Note: Automatic detection and classification of EEG artifacts using fuzzy kernel SVM and wavelet ICA (WICA)

    by K. Yasoda, R. S. Ponmagal, K. S. Bhuvaneshwari, K. Venkatachalam

  • Open Access

    Periodic Service Behavior Strain Analysis-Based Intrusion Detection in Cloud

    by S. Priya, R. S. Ponmagal

    The problem of intrusion detection in cloud environments has been well studied. The presence of adversaries would challenge data security in the cloud by generating intrusion attacks towards the cloud data and should be mitigated for the development of the cloud environment. In mitigating intrusion attacks, there exist several techniques in the literature. The method uses different features like frequency of access, payload details, protocol mapping, etc. However, the methods need to improve to achieve the expected performance in detecting intrusion attacks. An efficient Periodic Service Behavior Strain Analysis (PSBSA) is presented to handle this issue. Unlike earlier methods, the PSBSA model analyzes the behavior of users in various time frames like historical, recent, and current spans. The model focused on identifying intrusion attacks in several constraints, not just considering the current nature. The performance of intrusion detection can be improved by viewing the user's behavior in historical, present, and recent timespan. Unlike other approaches, the proposed PSBSA model considers the user's behavior at different times in measuring the user's trust towards intrusion detection. Accordingly, the proposed PSBSA model analyzes the behavior of users under various situations. It examines the behavior in accessing the services at historical, current, and recent times. The method performs Historical Strain Analysis (HSA) Current Strain Analysis (CSA) and Recent Strain Analysis (RSA). HSA analysis is performed according to the historical data, CSA is performed based on the current access data and RSA is performed with the recent access data. The model estimates various legitimacy support values on each analysis to conclude the trust of any user. According to the support values, intrusion detection has been performed. The proposed PSBSA model introduces higher accuracy in intrusion detection in a cloud environment.

  • Study on federated learning model with pruning, optimal quantization and selective updating for IOT and big data analytics

    by Ch. Ellaji, R. S. Ponmagal, V. Saritha

    The steady increase in various number of IoT devices continues to produce big data, but now the cloud centric methodology to IoT and Big data Analytics has beco The steady increase in various number of IoT devices continues to produce big data, but now the cloud centric methodology to IoT and Big data Analytics has beco