publications

2024

2023

1. Concurr Comput

   Support Vector Machine based spectrum handoff scheme for seamless handover in Cognitive Radio Networks

   Srikrishna Iyer, and Velmurugan T

   Concurrency and Computation: Practice and Experience,, 2023

   Abs Bib Code

   The advancements in wireless communication go in leaps and bounds ushering in due attention to spectrum sharing. Spectrum scarcity is one of the major limitations causing hardships in the existing wireless networks. Cognitive Radio Networks (CRNs) emerge as a solution to tide over such humps. It prompts the secondary user (SU) to look out for unused spectrum and utilize them. The CRN helps the SU by permitting it to switch over to unused portions of the spectrum. When a primary user (PU) claims back the spectrum, SU is obliged to perform a spectrum handoff. The SU decides the type of policy to be chosen for the handoff. Such a decision-making step during the handoff of the spectrum is imperative only if a changing policy is required. In this research work, Artificial Neural Networks (ANNs), Logistic Regression and Support Vector Machine (SVM) are proposed and implemented for a seamless handoff in CRN. From the experimental verifications, it is observed that the training accuracy is 97.9% and 97.6% for ANN and SVM, respectively. But during the actual phase, SVM to a certain extent performed better. This is due to the convergence nature of SVM on global minima.

     title = {Support Vector Machine based spectrum handoff scheme for seamless handover in Cognitive Radio Networks},
     author = {Iyer, Srikrishna and T, Velmurugan},
     journal = {Concurrency and Computation: Practice and Experience,},
     year = {2023},
     pages = {7534},
     paper = {https://onlinelibrary.wiley.com/doi/abs/10.1002/cpe.7534}
   }

2. arxiv

   GAT-GAN : A Graph-Attention-based Time-Series Generative Adversarial Network

   Srikrishna Iyer, and Teng Teck Hou

   2023

   Abs Bib Code

   Generative Adversarial Networks (GANs) have proven to be a powerful tool for generating realistic synthetic data. However, traditional GANs often struggle to capture complex relationships between features which results in generation of unrealistic multivariate time-series data. In this paper, we propose a Graph-Attention-based Generative Adversarial Network (GAT-GAN) that explicitly includes two graph-attention layers, one that learns temporal dependencies while the other captures spatial relationships. Unlike RNN-based GANs that struggle with modeling long sequences of data points, GAT-GAN generates long time-series data of high fidelity using an adversarially trained autoencoder architecture. Our empirical evaluations, using a variety of real-time-series datasets, show that our framework consistently outperforms state-of-the-art benchmarks based on \emphFrechet Transformer distance and \emphPredictive score, that characterizes (\emphFidelity, Diversity) and \emphpredictive performance respectively. Moreover, we introduce a Frechet Inception distance-like (FID) metric for time-series data called Frechet Transformer distance (FTD) score (lower is better), to evaluate the quality and variety of generated data. We also found that low FTD scores correspond to the best-performing downstream predictive experiments. Hence, FTD scores can be used as a standardized metric to evaluate synthetic time-series data.

   @misc{iyer2023gatgan,
     title = {GAT-GAN : A Graph-Attention-based Time-Series Generative Adversarial Network},
     author = {Iyer, Srikrishna and Hou, Teng Teck},
     year = {2023},
   }

2022

1. Sensors

   mm-Wave radar-based vital signs monitoring and arrhythmia detection using machine learning

   Srikrishna Iyer, and Muhammad Faeyz Karim

   Sensors,, 2022

   Abs Bib Code

   A non-contact, non-invasive monitoring system to measure and estimate the heart and breathing rate of humans using a frequency-modulated continuous wave (FMCW) mm-wave radar at 77 GHz is presented. A novel diagnostic system is proposed which extracts heartbeat phase signals from the FMCW radar (reconstructed using Fourier series analysis) to test a three-layer artificial neural network model to predict the presence of arrhythmia in individuals. The effect of person orientation, distance of measurement and movement was analyzed with respect to a reference device based on statistical measures that include number of outliers, mean, mean squared error (MSE), mean absolute error (MAE), median absolute error (medAE), skewness, standard deviation (SD) and R-squared values. The individual oriented in front of the radar outperformed almost all other orientations for most distances with an expected d = 90 cm and d = 120 cm. Furthermore, it was found that the heart rate that was measured while walking and the breathing rate which was measured for a motionless individual generated results with the lowest SD and MSE. An artificial neural network (ANN) was trained using the MIT-BIH database with a training accuracy of 93.9 % and an R2 value = 0.876. The diagnostic tool was tested on 15 subjects and achieved a mean test accuracy of 75%.

   @article{s22093106,
     title = {mm-Wave radar-based vital signs monitoring and arrhythmia detection using machine learning},
     author = {Iyer, Srikrishna and Karim, Muhammad Faeyz},
     journal = {Sensors,},
     year = {2022},
     pages = {3106},
     paper = {https://www.mdpi.com/1424-8220/22/9/3106},
   }

2021

1. Neural. Comput. Appl.

   Structural health monitoring of railway tracks using IoT-based multi-robot system

   Srikrishna Iyer, Velmurugan T, and Amir Gandomi

   Neural computing and applications,, 2021

   Abs Bib Code

   A multi-robot-based fault detection system for railway tracks is proposed to eliminate manual human visual inspection. A hardware prototype is designed to implement a master–slave robot mechanism capable of detecting rail surface defects, which include cracks, squats, corrugations, and rust. The system incorporates ultrasonic sensor inputs coupled with image processing using OpenCV and deep learning algorithms to classify the surface faults detected. The proposed Convolutional Neural Network (CNN) model fared better compared to the Artificial Neural Network (ANN), random forest, and Support Vector Machine (SVM) algorithms based on accuracy, R-squared value, F1 score, and Mean-Squared Error (MSE). To eliminate manual inspection, the location and status of the fault can be conveyed to a central location enabling immediate attention by utilizing GSM, GPS, and cloud storage-based technologies. The system is extended to a multi-robot framework designed to optimize energy utilization, increase the lifetime of individual robots, and improve the overall network throughput. Thus, the Low Energy Adaptive Clustering Hierarchy (LEACH) protocol is simulated using 100 robot nodes, and the corresponding performance metrics are obtained.

   @article{iyer2021structural,
     title = {Structural health monitoring of railway tracks using IoT-based multi-robot system},
     author = {Iyer, Srikrishna and T, Velmurugan and Gandomi, Amir},
     journal = {Neural computing and applications,},
     year = {2021},
     pages = {5897--5915},
     paper = {https://link.springer.com/article/10.1007/s00521-020-05366-9}
   }

2019

1. i-PACT

   Antlion optimization and whale optimization algorithm for multilevel thresholding segmentation

   Srikrishna Iyer, AP Nadkarni, and TN Padmini

   In , 2019

   Abs Bib Code

   Multi-level image segmentation is a critical task in image processing that involves multiple threshold values. As the high computational cost of an exhaustive search is inefficient and cumbersome, the optimal thresholds algorithms make for a better path to venture; hence a comparison of optimization algorithms to set the optimal thresholds is highly essential and beneficial. In this paper, a practical comparison is made to deduce the best optimization technique amongst the whale optimization and antlion optimization algorithm, to solve the multilevel threshold problem, to find the optimal multilevel thresholds. Otsu’s function is maximized to perform optimized thresholding-based image segmentation. The experimental results showed that the Antlion optimization algorithm gave better performance in solving the problem for higher level multi-thresholding.

   @inproceedings{8960178,
     title = {Antlion optimization and whale optimization algorithm for multilevel thresholding segmentation},
     author = {Iyer, Srikrishna and Nadkarni, AP and Padmini, TN},
     journal = {2019 Innovations in Power and Advanced Computing Technologies,},
     year = {2019},
     pages = {1-8},
     paper = {},
   }