Deep learning-enabled generative acceleration for topology-optimized structures in 2D and 3D domain S Kanmani, M. Murali Discover Artificial Intelligence, 2026 Latest research explored into how Deep Learning (DL) can be used to accelerate topology optimization while reducing processing costs. Topology optimization, relying on finite element analysis (FEA) and iterative solvers, typically incurs substantial computational overheads, especially in evaluating intricate designs. This paper investigates the effectiveness of various advanced neural network models in achieving faster topology optimization. We evaluate and compare three such models using three well-structured datasets. Our method produces topology-optimized designs in both the two- and three-dimensional domains, illustrating the efficacy of DL in this domain. According to the findings, both the upgraded U-Net and Res-U-Net systems perform well as dependable DL strategies and dependable as faster topology optimization. Notably, the analysis reveals that Res-U-Net exhibits superior performance at higher iterations compared to U-Net. Furthermore, we show that, while preserving excellent accuracy, our suggested CNN method has a large time advantage over current state-of-the-art techniques for training.
Deep Learning for Fungus Classification Parth Rathi, Smit Vichare, S. Kanmani Proceedings 2024 4th International Conference on Pervasive Computing and Social Networking Icpcsn 2024, 2024 This research work presents a novel approach for the classification of fungal species using ConvMixer, a recently proposed architecture that combines convolutional and mixer layers. Fungal infections pose a significant threat to agriculture and public health, making accurate and efficient classification methods crucial for disease management. The proposed method leverages ConvMixer blocks to extract hierarchical features from fungal images, allowing for robust classification across multiple species. This study has then trained and evaluated the proposed model on a dataset consisting of images from five different fungal species, achieving promising results with an average accuracy of 95 percent. Additionally, this study has conducted a comprehensive analysis of the proposed model’s performance, including confusion matrix visualization and random sample predictions, to assess its generalization capabilities. Overall, the research findings demonstrate the effectiveness of ConvMixer in fungal species classification and highlight its potential for real-world applications in agriculture and healthcare.
Predictive Analysis of Vehicle CO2-Emissions Samanyu B Rao, Nivedita Anand, S. Kanmani Proceedings 2024 4th International Conference on Pervasive Computing and Social Networking Icpcsn 2024, 2024 This research work presents a web application developed using the Flask framework for predicting and comparing the fuel consumption and C02 emissions of various vehicle models. The application leverages machine learning models, including linear model, ridge model, lasso model, elastic net model, neural network model, XG-Boost model and Random Forest model, to estimate fuel consumption and C02 emissions based on user-provided input features. The models are trained and loaded into the application, allowing users to select a vehicle and input relevant features for prediction. The proposed system identifies the best-performing model for each prediction, highlighting the closest prediction and its associated error percentage. Additionally, the application offers a comparison feature that enables users to compare the specifications of different vehicle models within the dataset. Users can select two vehicle models, and the system retrieves and displays their specifications, facilitating informed decision-making for consumers and researchers interested in understanding the environmental impact of vehicle choices. The web application provides an intuitive interface for exploring fuel consumption and emissions data, making it a valuable tool for both consumers and researchers in the automotive industry.
Price of Anarchy and Price of Stability Mapping for Analyzing Topology Design of Communication Networks Kanmani S., M. Murali Journal of Computer Science, 2023 The goal of game theory is to model actions among players or users in a common space who deal with various situations and face various outcomes. The study of game theory is widely applied to a wide range of economic fields, including auctions, renewable energy, wireless sensor networks, and software defined networks. Resource allocation and cooperation between networks or terminals are important in the field of game theory networking. In order to infer concrete solutions for the players, game formulas are used. A solution is determined by classifying players and calculating the Price of Anarchy (PoA) and Price of Stability (PoS) in order to determine Nash equilibrium and evaluate efficiency. Using the Open Systems Interconnection (OSI) layer as a lens, this study examines a variety of applications of game theory in non-cooperative environments and communication systems. This study focuses on the 'presence of governing' or participation nodes in a set of players in a network. Also, a comparison of different research fields in game theory is made.
Proposing a Hybrid Topological Organization for Non-Misbehaving Nodes with Optimal Path Selection Using Game-Theoretic Approach Kanmani S, M. Murali Journal of Computer Science, 2023 : In dynamic communication networks, improvement of energy efficiency is one of the major challenges for reliable communication. By considering this challenge, we focus to develop the topology of the network. In this study, we design a hybrid star-mesh topology for minimizing the latency as well as energy consumption of the network. In the hybrid network topology, the Ad-Hoc On-Demand Multipath Distance Vector (AOMDV) routing protocol is used to establish multiple paths between source and destination. Among the multiple paths, the optimal path is chosen using Chimp Optimization Algorithm (ChOA) when the routing path loses its energy level. The optimal path selection leads to enhancing the energy efficiency of the network. Simulation results discuss the superior performance of the proposed scheme in terms of delivery ratio, energy consumption, delay, and throughput by 7% on aggregate.
Computation of PoA for Selfish Node Detection and Resource Allocation Using Game Theory S. Kanmani, M. Murali Computer Systems Science and Engineering, 2023 The introduction of new technologies has increased communication network coverage and the number of associating nodes in dynamic communication networks (DCN). As the network has the characteristics like decentralized and dynamic, few nodes in the network may not associate with other nodes. These uncooperative nodes also known as selfish nodes corrupt the performance of the cooperative nodes. Namely, the nodes cause congestion, high delay, security concerns, and resource depletion. This study presents an effective selfish node detection method to address these problems. The Price of Anarchy (PoA) and the Price of Stability (PoS) in Game Theory with the Presence of Nash Equilibrium (NE) are discussed for the Selfish Node Detection. This is a novel experiment to detect selfish nodes in a network using PoA. Moreover, the least response dynamic-based Capacitated Selfish Resource Allocation (CSRA) game is introduced to improve resource usage among the nodes. The suggested strategy is simulated using the Solar Winds simulator, and the simulation results show that, when compared to earlier methods, the new scheme offers promising performance in terms of delivery rate, delay, and throughput.
Evaluating system cost and complexity of finding Nash equilibrium for non-misbehaving nodes in CSRA game Kanmani S, M. Murali Proceedings 4th International Conference on Smart Systems and Inventive Technology Icssit 2022, 2022 As the dynamic communication network's infrastructure is decentralised, certain nodes may act selfishly. It compromises the communication network's security and causes path delays, congestion, and inefficient resource utilisation, among other things. As a result, a Nash equilibrium-based game theory with the parameter Price of Anarchy was proposed to find selfish nodes (PoA). Non-cooperative or selfish nodes are identified using the parameter Price of Anarchy (PoA). Resources are allocated to the active or normal nodes after the selfish nodes have been detected (non-misbehaving nodes). The suggested resource allocation outperforms the prior work in terms of throughput, delay, and delivery ratio, according to the findings, which were implemented in Python and tested in the SolarWinds simulation environment.
Deep learning-enabled generative acceleration for topology-optimized structures in 2D and 3D domain S Kanmani, M Murali Discover Artificial Intelligence , 2026 2026
Employing an optimisation algorithm to design an evolutionary topological structure for non-misbehaving nodes with optimal path selection S Kanmani, M Murali Wireless Personal Communications 137 (3), 1423-1442 , 2024 2024
Skin disease detection S Kanmani, A Yaduka, P Verma AIP Conference Proceedings 3075 (1), 020097 , 2024 2024
Deep Learning for Fungus Classification P Rathi, S Vichare, S Kanmani 2024 4th International Conference on Pervasive Computing and Social … , 2024 2024
Predictive Analysis of Vehicle CO2-Emissions SB Rao, N Anand, S Kanmani 2024 4th International Conference on Pervasive Computing and Social … , 2024 2024
Computation of PoA for Selfish Node Detection and Resource Allocation Using Game Theory. S Kanmani, M Murali Computer Systems Science & Engineering 47 (2) , 2023 2023 Citations: 1
Computation and Analysis of Price of Anarchy for Selfish Node Detection and efficient Resource Allocation for Dynamic Communication Network using Game theory S Kanmani, M Murali 2022
Analyzing game theory applications in a layered perspective for a non-cooperative environment with the existence of nash equilibria in various fields of research S Kanmani, M Murali Proceedings of Third International Conference on Communication, Computing … , 2022 2022 Citations: 1
Evaluating system cost and complexity of finding nash equilibrium for non-misbehaving nodes in csra game S Kanmani, M Murali 2022 4th International Conference on Smart Systems and Inventive Technology … , 2022 2022 Citations: 1
MOST CITED SCHOLAR PUBLICATIONS
Computation of PoA for Selfish Node Detection and Resource Allocation Using Game Theory. S Kanmani, M Murali Computer Systems Science & Engineering 47 (2) , 2023 2023 Citations: 1
Analyzing game theory applications in a layered perspective for a non-cooperative environment with the existence of nash equilibria in various fields of research S Kanmani, M Murali Proceedings of Third International Conference on Communication, Computing … , 2022 2022 Citations: 1
Evaluating system cost and complexity of finding nash equilibrium for non-misbehaving nodes in csra game S Kanmani, M Murali 2022 4th International Conference on Smart Systems and Inventive Technology … , 2022 2022 Citations: 1
Deep learning-enabled generative acceleration for topology-optimized structures in 2D and 3D domain S Kanmani, M Murali Discover Artificial Intelligence , 2026 2026
Employing an optimisation algorithm to design an evolutionary topological structure for non-misbehaving nodes with optimal path selection S Kanmani, M Murali Wireless Personal Communications 137 (3), 1423-1442 , 2024 2024
Skin disease detection S Kanmani, A Yaduka, P Verma AIP Conference Proceedings 3075 (1), 020097 , 2024 2024
Deep Learning for Fungus Classification P Rathi, S Vichare, S Kanmani 2024 4th International Conference on Pervasive Computing and Social … , 2024 2024
Predictive Analysis of Vehicle CO2-Emissions SB Rao, N Anand, S Kanmani 2024 4th International Conference on Pervasive Computing and Social … , 2024 2024
Computation and Analysis of Price of Anarchy for Selfish Node Detection and efficient Resource Allocation for Dynamic Communication Network using Game theory S Kanmani, M Murali 2022
