Shanmugam Poonkuntran

The Satellite-Terrestrial Integrated Networks (STIN) were emerged as a key architectural pattern for attaining seamless, global and resilient wireless connectivity by adding extensive coverage of satellite systems with the high capacity and low latency of terrestrial networks. In spite of their advantages, STINs face significant challenges arising from heterogeneous link qualities, dynamic network topologies, long propagation delay and highly variable channel conditions which may complicate reliable and adaptable communication. The accurate and timely assessment of link quality is essential to enable effective resource management, adaptive modulation and coding and robust network control in space-ground integrated environments. In this study, AI-based link quality classification model for STINs based on a stacked ensemble learning architecture. This model combines multiple lightweight machine learning classifiers and a meta-level learner to capture complex non-linear relationships among satellite orbital parameters, spatial characteristics, and link dynamics. The framework categorizes satellite-terrestrial links into three operational states as Good, Moderate and Poor, which provides actionable intelligence for cross-layer resource allocation and adaptive communication strategies. The extensive experimental evaluation demonstrates that the proposed work attains 96.79% classification accuracy and 96.88% macro averaged F1-score where Decision tree as 85% and Machine Learning Based attain 88%. This indicates highly balanced and robust performance across all link classes. The confusion matrix analysis reveals that the misclassification occurs only between adjacent link quality states with no critical misclassifications between good and poor links. This ensures high reliability for operational decision-making. When compared to single-model baselines, the proposed approach increases prediction stability and robustness under heterogeneous and dynamic STIN conditions. The results confirmed the machine learning-assisted link quality intelligence can serve as a practical and efficient enabler for dynamic resource management in STIN. The model is computationally effective, scalable and readily deployable within next-generation STIN control planes which supports reliable communication for broadband access, emergency services and IoT applications.

In recent days, the rapid growth of advanced communication links among smart consumer appliances has enabled the integration of intelligent energy management system which combines sensing, communication, computing and control technologies within consumer electronics. A synthetic IoT sensor stream is developed to simulate real-time electricity consumption data from multiple utilities, transmitting sensor packets over a wireless uplink with realistic network conditions including jitter and packet loss. The main objective of this study is to design a lightweight hybrid framework which ensures convergence between cloud-level learning and edge-level control while maintaining bounded prediction error and latency. The convergence of the distilled-tree inference model with the cloud-based XGBoost model is achieved through minimization of distillation loss, guaranteeing stable edge predictions with steady-state variance below 3% across iterative updates. The proposed model ensures lightweight inference while the gradient boosting contributes to high predictive accuracy. The experimental results shows that the proposed model achieves about 17.6% simulated energy savings compared to traditional approaches like feedforward neural networks (12%) and random forest (10.3%). In addition, this study includes a prototype for consumer dashboard which showcases real-time device integration and futuristic energy management.

In the contemporary digital landscape, securing healthcare data stored on mobile devices has become imperative due to the increasing prevalence of cyberattacks. Healthcare data, often vulnerable to breaches, requires robust privacy-preserving mechanisms to ensure secure storage and sharing across various platforms. This study proposes a novel approach integrating TinyML (machine learning) with federated learning (FL) and differential privacy (DP) to enhance the security and privacy of healthcare data on resource-constrained mobile devices. The proposed TinyML model processes patient data, including ECG readings and cardiac arrhythmia diagnoses, directly on handheld devices, enabling real-time analysis with minimal resource consumption. Federated learning facilitates decentralized training of models on local devices, ensuring that sensitive data remains on user devices while model parameters are aggregated centrally. Differential privacy further strengthens security by introducing controlled noise into the data, safeguarding against malicious attacks without compromising data utility. The system demonstrates a high accuracy of 99%, significantly outperforming traditional models such as Decision Tree (76.0%), Random Forest (98.1%), and Federated Learning with Deep Q (95%). The proposed framework offers a scalable and efficient solution for real-time healthcare data processing, ensuring data privacy and security across distributed environments. Experimental results validate the model’s effectiveness in safeguarding healthcare data, with implications for broader applications in secure, privacy-preserving medical data analysis.

IoT (Internet of Things) devices are becoming more popular because they can be connected to mobile phones and communicate with various types of devices (e.g., smart homes, automation, and monitoring applications), which provides a high-quality service with the flexibility of performing operations from anywhere at any time. This research proposes a holographic display in consumer electronics to provide a more interactive experience for the user by providing significant security and privacy framework to the IoT devices which are prone to vulnerabilities. A simple Post-Quantum Cryptography Algorithm (PQCA) is used in this study to find flaws in a quantum lattice that is based on a 3x3 matrix in pre-processing. Also, an edge-based anomaly detection framework is created to work with consumer devices to show holographic data streams and IoT device activity in real time for devices with limited resources and less computational overhead than traditional machine learning models. The suggested method conducts testing on a set of unsupervised IoT devices, which contain data and metrics on user behaviour. The dataset contains 8000 service requests in an IoT environment. It finds security anomalies in 70% of the cases in training phases and 30% of the cases in testing phases, respectively. The proposed system detects vulnerable IoT devices by projecting them in a holographic projection graph, which in turn enhances privacy. The test results showed that the minimum encryption time overhead is 0.00085, and the accuracy achieved is about 97.44%.

In the medical field, medical imaging is essential for precise diagnosis, treatment planning, and condition monitoring. The goal of this work is to improve the field of healthcare imaging by investigating the complementary applications of augmented reality (AR) and artificial intelligence (AI) in semantic segmentation. More precise identification and delineation of anatomical structures and abnormalities in medical images is made possible by AI-driven semantic segmentation, opening the door to more precise diagnosis and treatment plans. Modern AI algorithms are employed in the suggested methodology to perform semantic segmentation in a variety of medical imaging modalities, including ultrasound, CT, and MRI scans. The detection of anomalies such as tumours, irregularities in organs, and neurological disorders is made easier by these algorithms. The foundation for integrating augmented reality technologies into the healthcare ecosystem is provided by the segmented medical images. AR enhances the interaction and visualization of segmented medical data in the context of healthcare. Real-time augmented reality overlays help surgeons by improving surgical navigation and precision. Additionally, AR apps help with medical education by offering professionals and students alike an immersive learning environment. AR provides interactive visualizations to help patients understand their medical conditions when they are going for therapy and rehabilitation. The difficulties in integrating these technologies in the healthcare industry are discussed in this paper, with a focus on the significance of data privacy, regulatory compliance, and easy integration with current healthcare systems. The successful deployment of AI technologies necessitates interdisciplinary collaboration among AI developers, healthcare professionals, and AR specialists to ensure compliance with ethical, legal, and clinical standards mandated in the healthcare domain. The proposed Segmentation mask is redefined with geometric calibration and used with U-Net for image segmentation. The segmentation is experimented on Cityscapes and PASCAL VOC 2012 datasets. The experimental results show that proposed semantic segmentation based on geometric calibration yields more accuracy than its counterparts.

Due to advancements in privacy, integrity and confidentiality have become increasingly important when transferring data over the Internet. Without proper security measures, private data could be viewed by outsiders or attackers during transmission. To address this issue, two different methods can be implemented to increase information security, one is face shape measurement, and another one is secure medical image processing. This article discusses how deep learning techniques can be combined with secure healthcare systems to improve safety and provide a consistent and flawless system. The process begins with capturing and analyzing a person’s face, generating facial features using deep learning. The separated facial features are then converted into a binary stream to form a deep learning interface, which is discarded for another cryptanalytic process (encrypting and deciphering) and stored as a pattern in the database. This warehoused value is used as a key to encrypt data, which is ready to be communicated over the web. On the sender’s side, the data are converted into encrypted data using their public key, their private key, and the receiver’s and sender’s public keys. Upon receiving the data, the recipient uses their private key to decrypt and access the data. With this approach, private data are well protected as the facial recognition of the person is used as a key to secure the health data from criminals. Samples are converted to unique keys generated by doing this. Converting samples to original keys generated by a 97% accurate face extraction service in everyone so that their players do not have access to private information and to improve information privacy and security is achieved.

Crowd behavior prediction (CBP) and crowd counting (CC) are the essential functions of vision-based crowd analysis (CA), which play a crucial role in controlling crowd disasters. The CA using different models for the CBP and the CC will increase computational overheads and have synchronization issues. The state-of-the-art approaches utilized deep convolutional architectures to exploit spatial-temporal features to accomplish the objective, but such models suffer from computational complexities during convolution operations. Thus, to sort out the issues as mentioned earlier, this paper develops a single deep model which performs two functionalities of CA: CBP and CC. The proposed model uses multilayers of depth-wise separable CNN (DSCNN) to extract fine-grained spatial-temporal features from the scene. The DSCNN can minimize the number of matrix multiplications during convolution operation compared to traditional CNN. Further, the existing datasets are available to accomplish the single functionality of CA. In contrast, the proposed model needs a dual-tasking CA dataset which should provide the ground-truth labels for CBP and CC. Thus, a dual functionality CA dataset is prepared using a benchmark crowd behavior dataset, i.e. MED. Around 41 000 frames have been manually annotated to obtain ground-truth crowd count values. This paper also demonstrates an experiment on the proposed multi-functional dataset and outperforms the state-of-the-art methods regarding several performance metrics. In addition, the proposed model processes each test frame at 3.40 milliseconds, and thus is easily applicable in real-time.

This chapter presents an integer transform-based watermarking scheme for digital fundus image authentication. It is presented under multimedia applications in medicine. The chapter introduces image authentication by watermarking and digital fundus image. The key requirements in developing watermarking scheme for fundus images and its challenges are identified and highlighted. Authors describe a proposed watermarking scheme on integer transform. The experimental results emphasize the proposed scheme's ability in addressing key requirements and its attainment. The detailed results are summarized.