Afsana Ahamed

Dr. Afsana Ahamed

Arkansas Tech University, United States
+1
orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 May 2018
Verified
Conference Paper
Closed access
AAAfsana Ahamed
Afsana Ahamed
HVHamid Vakilzadian
Hamid Vakilzadian
Intelligent transportation systems (ITS) are becoming more popular due to the demand for advanced cyber-physical systems and comfort applications and services needed for use in autonomous vehicles. Vehicular ad hoc Networks (VANETs) are an essential part of ITS, as they support vehicle-to-vehicle and vehicle-to-infrastructure data communication. The performance of a VANET solely depends on the routing protocols in the network layer, which provide efficient and effective data communication. In this paper, we survey the most recent issues surrounding efficient routing protocols that are suitable for VANETs. We also address the pros and cons of these protocols in terms of scalability, quality of services, security and privacy, energy efficiency, transmission bandwidth limitation, and broadcasting issues.

Verified
Journal Article
Open access
PDF file
AAAfsana Ahamed
Afsana Ahamed
HVHamid Vakilzadian
Hamid Vakilzadian
A vehicular ad hoc network (VANET) is a technology in which moving cars are used as routers (nodes) to establish a reliable mobile communication network among the vehicles. Some of the drawbacks of the routing protocol, Ad hoc On-Demand Distance Vector (AODV), associated with VANETs are the end-to-end delay and packet loss. We modified the AODV routing protocols to reduce the number of route request (RREQ) and route reply (RREP) messages by adding direction parameters and two-step filtering. The two-step filtering process reduces the number of RREQ and RREP packets, reduces the packet overhead, and helps to select the stable route. In this study, we show the impact of the direction parameter in reducing the end-to-end delay and the packet loss in AODV. The simulation results show a 1.4% reduction in packet loss, an 11% reduction in the end-to-end delay, and an increase in throughput.

23 November 2023
Verified
Conference Paper
Closed access
AAAfsana Ahamed
Afsana Ahamed
AHA. Hamdhee
A. Hamdhee
MAM.N.M. Aashiq
M.N.M. Aashiq
RHR. Hirshan
R. Hirshan
WKW.G.C.W. Kumara
W.G.C.W. Kumara
Accurately diagnosing the stages of Alzheimer's disease (AD) using MRI images poses a challenge for medical practitioners, as even experienced professionals can make mistakes in diagnosing the stage. Traditional solutions heavily rely on various MRI pre-processing techniques to improve diagnostic accuracy. In recent years, deep learning methods, particularly Convolutional Neural Networks (CNNs), have shown promise in AD diagnosis through classification approaches. However, there is still a need to bridge the research gap in utilizing deep learning techniques for object detection in Alzheimer's disease diagnosis. In this research, comprehensive comparisons are conducted and various techniques to address this gap are explored, aiming to identify the most effective approach for solving this diagnostic challenge. We introduce hyperparameter tuning in our CNN model to identify the best parameters for achieving optimal accuracy by considering execution time. Next, we compare the accuracy and loss metrics of various CNN architectures' respective execution times. Additionally, we examine the impact of dataset augmentation techniques and the number of data samples on accuracy, loss, and execution times. Our proposed CNN model incorporates the latest techniques, featuring the EfficientNetV2S architecture. Hence, our work provides a benchmark for deep learning-based Alzheimer's disease stage detection, categorizing individuals into four stages Mild Demented, Moderate Demented, Non Demented, and Very Mild Demented. The utilization of the dataset using MRI data from the ADNI, with its extensive 2D slices for each stage, enables the extraction of meaningful features and patterns associated with Alzheimer's disease pathology. The outcomes of this study contribute to the advancement of diagnostic tools and therapeutic interventions, enhancing the understanding and management of Alzheimer's disease.

30 August 2026 in Energies
Verified
Journal Article
Open access
PDF file
DWDillon Michael Wood
Dillon Michael Wood
BLBenjamin Leon
Benjamin Leon
EMEthan Mashburn
Ethan Mashburn
AAAfsana Ahamed
Afsana Ahamed
Seyed Ehsan HosseiniSeyed Ehsan Hosseini
Seyed Ehsan Hosseini
To encourage the use of hydrogen energy and improve the dependability of hydrogen fuel cells, this study develops a two-stage semi-supervised framework for anomaly pattern discovery and codification. Isolation Forest, One-Class Support Vector Machine, and Local Outlier Factor are first applied to unlabeled time series data; six supervised classifiers are subsequently trained on ensemble-derived pseudo-labels to learn these patterns efficiently. The analysis uses a reproducible 20% sample (random seed 42) of 185,721 observations from four fuel cells in the NASA Prognostics Data Repository, yielding 37,144 observations. The three unsupervised algorithms exhibit distinct detection patterns, with 280 common detections (15.07% of each model’s flagged observations and 0.754% of the sample). A consensus-weighted ensemble identifies 1296 potential anomalies (3.49%). Random Forest best reproduces the ensemble pseudo-labels, with 99.365% testing accuracy; this value measures pattern learnability and is not independent verification of physical faults. Gini importance identifies power difference, capacity, load power, load current, measured power, and load resistance as the principal predictive variables. Because independently confirmed fault labels and maintenance outcomes are unavailable, the reported detections are interpreted as systematic deviations requiring electrochemical or expert verification.