Computer science researcher with strong problem-solving skills. Working on state-of-the-art technologies related to image forensics, artificial intelligence, and deep learning. A team player with practical knowledge of multi-cultural & diverse teams, seeking a career in research & academics with 10+ years experience in academics and research.
RESEARCH, TEACHING, or OTHER INTERESTS
Computer Vision and Pattern Recognition, Computational Theory and Mathematics, Artificial Intelligence
EVA-centric QUBO optimization for active and reactive power coordination in DLMP-driven distribution systems Vansh Suri, Neelu Nagpal, Pierluigi Siano, Ravi Sharma, Saurabh Agarwal, Wooguil Pak Scientific Reports, 2026 This research presents a bilevel optimization framework enhanced with quadratic unconstrained binary optimization (QUBO) for the reactive demand response (DR) driven by the locational marginal price (DLMP) of electric vehicle aggregators (EVA) in renewable integrated distribution systems. The methodology fundamentally reverses the traditional power system hierarchy by positioning the EVA as the Stackelberg leader and the distribution system operator (DSO) as the follower, acknowledging the growing influence of electric vehicles as dominant flexible resources. The proposed framework introduces the first comprehensive approach that simultaneously optimizes both active power charging schedules and reactive power provision through QUBO-enhanced spectral clustering with multimetric validation. It also introduces a novel post-optimization disaggregation methodology that distributes cluster-level decisions to individual vehicles while minimizing charging interruptions. The comprehensive evaluation of the IEEE 33 bus distribution system with 1000 electric vehicles demonstrates quantifiable economic benefits that include a 24.70% reduction in DSO costs, a 75.93% reduction in EVA payments, and 39.55% total system cost savings compared to conventional approaches. The technical improvements include a 4.64% reduction in power loss, 9.86% improvement in voltage deviation, 6.49% improvement in the load factor, and 86.11% effectiveness of the voltage support. QUBO optimization achieves a 100% success rate with 72,000 binary variables while introducing reactive power revenue streams generating $47.20 per day, while maintaining 98.7% customer satisfaction. The quantum-inspired optimization approach positions the framework for future enhancement as quantum computing technologies mature, providing immediate benefits through algorithms capable of handling the computational complexity of large-scale EV integration scenarios. The methodology has been compared against Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), and the QUBO problem has been simulated using a Linprog, a solver in Yalmip toolbox in MATLAB, and compared against a commercial solver, Gurobi, for computation analysis. The results establish performance benchmarks for advanced EV integration strategies, providing validation data for regulatory frameworks and investment decisions in distribution systems with high penetration of electric vehicles and renewable energy.
Advanced security in fog environments using encryption and adaptive user activity tracking Hari Mohan Rai, Abdul Razaque, Neha Agarwal, Saurabh Agarwal, Hanaa A. Abdallah, Shashi Kant Scientific Reports, 2026 The use of fog computing is on the rise, adding new dimensions to security and, more specifically, to data protection in fog cloud environments. Storing fog-computing data increases the likelihood of data exploitation when it is uploaded to fog-computing storage. In this paper, Adaptable User Activity Tracking (ASUT) is introduced, integrating AES-256, SHA-512, and user activity tracking (UAT). The need to integrate activity monitoring into the ASUT to collect statistical information on user actions has been stated. The file uploaded to the fog computing storage is encrypted using a 256-bit AES key. Then, this key is hashed with SHA-512 and stored in the fog cloud. The AES expansion is used to decrypt the data, while the SHA-512 hash of the AES key is used to verify that the user-provided key matches the original before decryption proceeds-the hash is irreversible, and the original key is never stored in plaintext. The user must know the initial key to access the file further. When the client re-enters the fog, the algorithm compares the hashes of the two: the initial and the second entry keys. In parallel, the fog cloud broadcasts the user's actions to track any abnormal activity on the account. This mechanism helps mitigate risks of unauthorized data access and suggests ways to improve user protection. The proposed ASUT is designed using Python and PHP. Experimental results show that ASUT achieves 43.39% faster encryption, 66% faster decryption, and 19.86% higher throughput compared to the best-performing competing method, indicating improved computational efficiency and practical feasibility under the evaluated conditions.
GAN-TL domain-adaptive augmentation for enhanced transfer learning Jenefa Archpaul, Vidhya Kandasamy, Manoranjitham Rajendran, Thompson Stephan, Punitha Stephan, Saurabh Agarwal, Wooguil Pak Journal of Big Data, 2026 Transfer learning offers a practical solution for label-scarce target domains, but its effectiveness diminishes when source and target data follow different distributions. To address this challenge, we present GAN-TL, a GAN-augmented transfer learning framework that generates target-style samples to expand target-relevant variability and couples this synthesis with adversarial feature alignment. The task model is trained on mixed real and synthetic batches, enabling improved cross-domain generalization under limited target supervision. Experiments on three benchmarks (Datasets A–C) confirm consistent improvements over standard transfer learning. On Dataset A, GAN-TL achieves 87.4% accuracy compared with 83.5% for conventional fine-tuning, and increases macro-F1 from 83.4% to 87.4%. For the regression benchmark Dataset B, mean squared error decreases from 9.72 to 7.89 while $$R^2$$ improves from 0.83 to 0.88. On the unseen-domain Dataset C, GAN-TL raises accuracy to 82.3%, reduces domain adaptation loss from 0.25 to 0.19, and lowers generalization error from 0.22 to 0.18. Robustness evaluations further show improved resistance to adversarial perturbations, with FGSM attack success reduced to 39.2% (vs. 48.3%) and average DeepFool perturbation lowered to 0.009 (vs. 0.012). Despite the added GAN branch, training converges faster in epochs, reaching stability at 26 epochs with a final loss of 0.53 (vs. 29 epochs and 0.62 for standard transfer learning). Overall, the results demonstrate that target-style synthesis combined with principled alignment yields a stable and robust transfer pipeline under pronounced domain shift.
Blockchain-based two-level trustable reputation framework for e-commerce platform using smart contracts K. Sundara Krishnan, R. Chithra Devi, Christo Ananth, D. Easwaramoorthy, Saurabh Agarwal, Wooguil Pak, Hari Mohan Rai, Shashi Kant Scientific Reports, 2026 E-commerce platforms incorporate reviews and reputation systems, allowing retailers and customers to manage and track their financial transactions. Consequently, it is crucial to design a reliable reputation system for the e-commerce environment, as it faces well-documented threats, including sybil attacks, feedback collusion, impersonation, review tampering, and whitewashing attacks. Current centralized systems are vulnerable to impersonation attacks, feedback manipulation, and lack automated verification against collusion-based reputation distortion. These unwanted ratings and reviews are highly correlated with abnormal cyber-attacks that damage both seller reputations and buyer experiences. To address these challenges, we propose a Blockchain-based Two-Level E-Commerce Trustable Reputation Framework (BTL-ETRF) utilizing deep learning-embedded transformers and redactable blockchain systems. Initially, we implement Multi-Factor Authentication for e-commerce users, utilizing three factors: PIN, OTP, and biometric fingerprint, to mitigate impersonation attacks. Only authenticated users are allowed to proceed to the reputation verification stage, where the proposed work considers five major metrics to classify user reputation using the Residual Dilated Convolution Transformer. To automate the reputation verification process, we design and employ two smart contracts, the Authentication Smart Contract and the Reputation Smart Contract which trigger automated actions based on the BTL-ETRF results. All transactions include reputation classification, and triggered actions are stored in the redactable blockchain, which can modify the stored transactions if needed. Finally, we demonstrate the performance of the proposed BTL-ETRF using Python and Ethereum Solidity, and conduct a formal analysis that shows the proposed model outperforms the compared works.
Dynamic Simulation and Optimization of an Innovative Cogeneration System Using TRNSYS, EES, and Response Surface Methodology as a Machine learning method Ali Dezhdar, Ehsanolah Assareh, Jagadeesh Kumar Alagarasan, Mehdi Hosseinzadeh, Saurabh Agarwal, Saleh Mobayen, Neha Agarwal Journal of the Taiwan Institute of Chemical Engineers, 2026 The rapid growth in global energy demand, coupled with the urgent need to reduce greenhouse gas emissions, necessitates the development of sustainable, multi-output energy systems. This study presents the design, simulation, and optimization of an innovative solar-powered cogeneration system capable of simultaneously meeting the annual electricity, heating, cooling, and freshwater demands of a 100-unit residential complex in Tehran, Iran. The proposed configuration integrates a solar dish concentrator, organic Rankine cycle (ORC), heat pump, reverse osmosis desalination, hot and cold water storage tanks, and an auxiliary boiler. Dynamic simulations were conducted using TRNSYS, with thermodynamic modeling of the ORC performed in Engineering Equation Solver (EES). A multi-objective optimization framework, based on Response Surface Methodology (RSM) in Design Expert, was applied to minimize life cycle cost (LCC), boiler fuel consumption, and electricity usage, while maximizing thermal comfort (predicted mean vote, PMV).The optimal design—comprising a 101.25 m² solar dish, 20.27 kW cooling capacity, 36.29 kW heating capacity, and benzene as the ORC working fluid—achieves an LCC of $134,130, annual boiler fuel consumption of 6,918.0 m³, and electricity consumption of 6,437.4 kWh, while producing freshwater with a total dissolved solids (TDS) level below 500 mg/L in compliance with WHO drinking water standards. Compared with a conventional PV + grid + boiler system, the proposed design reduces the LCC by 18.4% and annual CO₂ emissions by over 13 tonnes. Although benzene emerged as optimal in the numerical optimization, alternative low-global-warming-potential fluids such as R1233zd(E) are discussed as safer, environmentally favorable options with minimal efficiency trade-offs. In particular, R1233zd(E) and toluene are highlighted as practical alternatives, balancing acceptable efficiency with improved safety and environmental compliance. Beyond the Tehran case study, a dimensionless scaling approach is proposed for adapting the system to other climatic zones, enabling re-optimization of key parameters such as solar dish area and heat pump capacities based on local meteorological and demand profiles. The results demonstrate that the integrated TRNSYS–EES–RSM framework can deliver cost-effective, environmentally sustainable, and technically robust cogeneration solutions for residential applications in sun-rich regions.
Quantum blockchain: Trends, technologies, and future directions Manjula Gandhi S, Chaitrali Mulay, Karthiganesh Durai, G. Murali, Jafar Ali Ibrahim Syed Masood, V. Vijayarajan, Kumar Gautam, N. S. Kalyan Chakravarthy, S. Suresh Kumar, Saurabh Agarwal, Murali S, Vijayasherly V, David Asirvatham, Sarfraz Brohi, Chandru Vignesh C, Anbuchelian S Iet Quantum Communication, 2024
