Univrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001A Trust-Aware and Fault-Tolerant Framework for Secure Data Sharing in Cloud–Edge Environments11710891210.22059/jitm.2026.108912ENKalpanaDwivediPh.D. Candidate, Department of Computer Science and Engineering, SSCSE, Sharda University, Greater Noida, Uttar Pradesh 201310.GauravRajAssociate Prof., Department of Computer Science and Engineering, SSCSE, Sharda University, Greater Noida, Uttar Pradesh 201310.Journal ArticleThe accelerated growth of cloud edge computing has increased the need for secure, reliable, and efficient data exchange in distributed and heterogeneous environments. However, node failures, untrustworthy edge devices, information corruption, key destruction, and excessive communication overhead are significant issues that significantly affect the performance and security of the system. To address these challenges, this paper proposes a trust-aware, fault-tolerant, and secure data-sharing framework for cloud–edge environments. The framework combines the use of dynamic trust evaluation, lightweight cryptography, and erasure codes for data distribution to ensure data confidentiality, integrity, availability, and trustworthiness. By a behavior-based trust model, the edge nodes are continually evaluated on parameters such as reliability, history of previous interactions, and consistency of services, allowing for intelligent decisions to be made regarding data placement and recovery. Erasure coding allows for reconstructing data against multiple node failures at a low storage overhead cost, which enables greater fault tolerance. Within the same framework, secure key management and integrity checks are provided against eavesdropping, replay attacks, insider attacks, and malicious node actions. Simulation results demonstrate that the proposed framework significantly outperforms conventional cloud–edge data-sharing solutions in terms of data availability, fault tolerance, latency, and bandwidth efficiency. The results demonstrate the suitability of the proposed framework for large-scale and security-sensitive applications, such as IoT data aggregation, healthcare, and enterprise cloud computing.https://jitm.ut.ac.ir/article_108912_f062af48cd10bc3a870303bc7c55a946.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001Towards Enhanced Biometric Systems: Insights into Human Gait Recognition184310891310.22059/jitm.2026.108913ENAppoorvaBansalDepartment of Computer Science and Information Technology Teerthanker Mahaveer University (TMU), Moradabad, Uttar Pradesh, India.RakeshDwivediProf., Department of Computer Science Engineering, Teerthanker Mahaveer University, Moradabad, India (TMU), Moradabad, Uttar Pradesh, India.ArpitJainProf., Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Guntur, A.P., India.Journal ArticleGait recognition is a biometric method of identifying people based on their walking pattern, which is the result of the coordinated action of the brain, nervous system, and muscles. Historically, gait analysis has been performed using a simple visual approach, but now several new tools have emerged, such as wearable sensors and vision-based motion capture, and a nonintrusive approach to facilitate gait analysis. In this paper, a detailed review of human gait recognition is presented. This is followed by a bibliometric analysis of the trends of publication, major sources, and country-wise research contributions. The current state of the art of research is then summarized, ranging from handcrafted feature-based approaches to recent deep learning architectures and the public datasets and benchmarks that have been used, from 2005 onwards, to demonstrate how performance has improved over time using the CASIA-B dataset. In addition to the technical aspects, the applications of gait recognition in surveillance, healthcare, sports, and defense are discussed. Last, we discuss some future directions, including cross-domain learning, transformer-based models, etc., and present key challenges, including domain shift, covariate sensitivity, data dependency, and privacy concerns. The collective discussions provide a framework for researchers and practitioners in gait recognition and identify the gaps that still need to be filled to enable more widespread adoption of gait recognition in practice.https://jitm.ut.ac.ir/article_108913_cd438fb2824eb980a6d8940c193890ba.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001A Novel Edge AI Framework for Selective Subtask Offloading in UAV-Assisted Multiuser Multiserver Mobile Edge Computing445810891410.22059/jitm.2026.108914ENSuhelAhamedDepartment of Information Technology, Guru Ghasidas Vishwavidyalaya (A Central University), Bilaspur, Chhattisgarh, India.Amit KumarDewanganDepartment of Information Technology, Guru Ghasidas Vishwavidyalaya (A Central University), Bilaspur, Chhattisgarh, India.AbhishekJainDepartment of Information Technology, Guru Ghasidas Vishwavidyalaya (A Central University), Bilaspur, Chhattisgarh, India.NishantBeharDepartment of Computer Science and Engineering, Guru Ghasidas Vishwavidyalaya (A Central University), Bilaspur, India.Pankaj ShankarShrivastavaDepartment of Electronics and Communication Engineering, Guru Ghasidas Vishwavidyalaya (A Central University), Bilaspur, India.Journal ArticleMobile edge computing (MEC) integrated with unmanned aerial vehicles (UAVs) enables low-latency, energy-efficient computation for resource-constrained ground users; however, selective subtask offloading in multiserver, multiuser environments remains a challenging combinatorial optimization problem. A novel IT management framework is needed to address the challenges of modern smart computing environments and support efficient, adaptive operational decision-making. This paper proposes an Edge AI framework for selective subtask offloading built on the public IEEE Dataport MSMU-CO dataset; we use the term oracle to denote the minimum-cost multi-commodity flow (MCMF) solution, which serves as an offline near-optimal benchmark for generating training labels and evaluating policies. Starting from 800 MCMF-generated multi-server, multiuser scenarios, each user task is decomposed into K=3subtasks, oracle labels are derived under an explicit latency-energy cost model, and the data are augmented with UAV-supported MEC geometry, including aerial and ground servers, 3D positions, line-of-sight (LoS) probability, and effective uplink rates. To ensure realistic deployment conditions, a deliberate feature leakage guard eliminates anchor-derived geometries, preventing data leakage from post-optimization target assignments and maintaining causal inference realism. A compact multilayer perceptron (MLP) is then trained to predict per-subtask offloading decisions solely from task, channel, and localized UAV context features. Evaluated on a test set containing 1,780 users and 5,340 subtasks, the proposed model achieves an accuracy of 0.842, precision of 0.793, recall of 0.870, an F1 score of 0.830, and an AUC of 0.926 under the cost model used for label generation. Furthermore, we benchmark our approach against a greedy linear relaxation baseline (LR Greedy) based on Huang et al. (2018, 2020). Experiments show that both the proposed policy and the greedy baseline significantly reduce deadline violations and mean system costs compared with an All-Local execution strategy, effectively bridging the performance gap toward the offline oracle labels in varying UAV-assisted MEC environments.https://jitm.ut.ac.ir/article_108914_ae0e1b9353d3f43727c6bd6b902c8499.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001Hybrid Secure Encryption-Based Mathematical Model for Machine Learning-Driven Cardiac Disease Prediction596910891510.22059/jitm.2026.108915ENPriyaGuptaResearch Scholar, Department of Computer Science and Engineering, Sanskriti University - Mathura-281401, India.SakuldeepSinghAssociate Professor, Department of Computer Science and Engineering, Sanskriti University- Mathura-281401, India.Journal ArticleDiseases related to the heart are among the most common causes of death worldwide. Therefore, it is essential to have a correct, reliable, and secure predictive model to support timely clinical decision-making. This study presents a machine learning-based hybrid mathematical model along with encryption for predicting heart diseases. This framework includes robust data preprocessing, feature encryption, and collective learning to improve diagnostic accuracy while ensuring the confidentiality of patients’ data. The framework uses three main classifiers—Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), and Decision Tree (DT)—along with a voting-based collaborative approach that combines their complementary strengths. In this study, a dataset of 1,026 cases with 14 medical parameters related to cardiac disease from a public dataset is used for the experimental setup. F1-score, recall, precision, and accuracy are considered performance evaluation metrics. The proposed framework achieves 98.10% accuracy, compared with 82.10% for Decision Tree, 90.90% for Support Vector Machine (SVM), and 94.50% for XGBoost, along with 98.12% precision, 98.00% F1-score, and 96.00% recall, demonstrating improved flexibility and generalization. The secure encryption layer further enhances the framework by protecting sensitive patient features while preserving statistical learning properties.https://jitm.ut.ac.ir/article_108915_8c01eaa44e3e785dc3fcba07c879aa6a.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001Transparent, Cost-Sensitive Threshold Calibration for Autonomous Edge-AI Health Wearables: A Severity-Stratified Hysteretic Algorithm for Predictive Asthma Facemasks709210891610.22059/jitm.2026.108916ENTarunTSchool of Computer Science and Engineering, Lovely Professional University, Phagwara, Punjab, India.Vijay KumarGargSchool of Computer Science and Engineering, Lovely Professional University, Phagwara, Punjab, India.VarshaSahniSchool of Computer Science and Engineering, Lovely Professional University, Phagwara, Punjab, India.VSharmaDepartment of Computer Science & Engineering, Graphic Era Hill University, Dehradun, India.SVatsComputer Vision Laboratory, Department of Computer Science & Engineering, Madan Mohan Malaviya University of Technology, Gorakhpur, Uttar Pradesh - 273010, India.Journal ArticleOrganizations deploying autonomous, AI-enabled health wearables must govern the opaque decision rules that gate their interventions, which carry direct consequences for cost, safety, and accountability. Wearable predictive devices for asthma combine environmental and physiological sensing with on-device actuation, yet these decision thresholds are almost always fixed heuristics. Fixed thresholds ignore inter-patient severity, trade sensitivity against false alarms suboptimally, and cause actuator chattering that wastes battery and metered medication. This paper presents AeroGate, a single-threshold calibration algorithm for a smart, AI-enabled asthma facemask that warms inhaled air, classifies cough acoustics, and triggers a vibrating-mesh nebulizer. AeroGate derives a population severity prior from chest imaging and converts a calibrated risk score into per-patient actuation thresholds through imaging-driven severity stratification, a cost-sensitive base threshold that weighs a missed attack against a false alarm, a data-sized hysteresis band that suppresses chattering, and an O(1) on-device update that holds a target alarm budget. Evaluated on the NIH ChestX-ray14 benchmark (112,120 studies) with strict patient-wise splits, the calibrated head attains an AUC of 0.68 and a Brier score of 0.23; AeroGate reduces actuator toggling by 77% relative to a single-threshold policy while matching detection sensitivity and holds the realised alarm rate at its target budget. Beyond the engineering, AeroGate reframes the operating point as a transparent, auditable management lever: the alarm budget becomes a cost-and-service target that managers can hold, and the cost ratio becomes a risk-tolerance input that clinicians and managers can set, yielding a principled, low-infrastructure recipe for deploying, governing, and scaling autonomous edge-AI health interventions.https://jitm.ut.ac.ir/article_108916_e8e59f2b4f3c9aa221f3cf1d6c11fddf.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001Explainable Privacy-Preserving Federated Learning for Collaborative Gout Risk Management Using Genetic and Lifestyle Data9311610891710.22059/jitm.2026.108917ENYu-RueyLiuCollege of Information and Electrical Engineering, Asia University, Taichung, Taiwan; Department of Emergency Medicine, Cheng Ching General Hospital, Taichung, Taiwan.Chun-YuanLinDepartment of Computer Science and Information Engineering, Asia University, Taichung, Taiwan.Yo-YuLiuDepartment of Business Administration, Nanhua University, Chiayi 622, Taiwan.SnehaPatnaikSchool of Public Health, KIIT-DU, Bhubaneswar, India.KhemrajSharmaSchool of Management, Kalinga Institute of Industrial Technology, Bhubaneswar, India.Journal ArticleBeyond improving predictive performance, the proposed framework contributes to healthcare information management by enabling secure collaboration among geographically distributed healthcare institutions while complying with data privacy regulations. The explainable prediction mechanism enhances clinicians' confidence in AI-assisted decision-making by identifying key genetic and lifestyle risk factors that influence preventive interventions. The proposed framework offers practical value for hospital administrators, healthcare policymakers, and digital health managers seeking scalable, privacy-preserving, and interpretable AI solutions for collaborative clinical analytics and precision healthcare.https://jitm.ut.ac.ir/article_108917_1f2c4c4aa2d36dd49a4cee8f253c0a9e.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001Deep Learning Iris Cryptography: A Deep Learning-Based Image Encryption Framework for Secure Information Management11713310891810.22059/jitm.2026.108918ENAwadhesh KumarSinghPh.D. Scholar, Department of CSE, Sharda School of Engineering & Technology, Sharda University, Greater Noida-201306 UP, India.Amrit KumarAgrawalAssociate Professor, Department of CSE, Sharda School of Engineering & Technology, Sharda University, Greater Noida-201306 UP, India.Journal ArticleAs high-resolution image data increasingly becomes commonplace in healthcare, surveillance systems, smart environments, and cloud platforms, ensuring image security has become one of the major challenges of our time. In this paper, we develop DL-IrisCrypt, a novel biometric-based image encryption framework in which the ciphertext is associated with the identity of the authorized user. We extract a 2048-bit iris code using Daugman segmentation and Gabor filters, which is then used to generate a key for initializing a six-dimensional hyperchaotic system and processed through SHA-512 and HKDF. To enhance randomness in the key-generation process, we incorporate features extracted from the latent space of a properly trained U-Net autoencoder model. The encryption architecture uses Fisher-Yates permutation, two layers operating with PCG64 in counter mode, and three passes of BLAKE2b cipher-block chaining. Experimental results show that the proposed approach achieves an NPCR of 99.61%, a UACI of 33.46%, and an information entropy of 7.9966 bits/byte, while passing all randomness tests provided by NIST SP 800-22. The framework requires an average of 16 milliseconds for encryption and 188 milliseconds for end-to-end processing on a single GPU and has an effective key space of 2^512. In addition to providing robust information security, DL-IrisCrypt offers features that can benefit organizations handling sensitive visual materials. By combining biometric identification, deep learning, and lightweight encryption mechanisms, the system enables secure management and preservation of the confidentiality of important digital assets while facilitating efficient information sharing with low computational requirements.https://jitm.ut.ac.ir/article_108918_4183f64f3ab1929ac1c244f8fe0b9ba5.pdfUnivrsity Of Tehran PressJournal of Information Technology Management2980-797218420261001A Smart Healthcare Management System for Predicting ICU Length of Stay Using Machine Learning and MIMIC-IV13414910891910.22059/jitm.2026.108919ENDanishAtherProfessor, Department of Computer Science and Engineering, Tashkent University of Information Technologies named after Muhammad al-Khwarizmi, Tashkent 100084, Uzbekistan.DharmRajAssistant Professor, Department of Computer Science and Engineering, Sharda School of Engineering & Technology, Sharda University, Greater Noida, Uttar Pradesh 201306, India.Suhail JavedQuraishiAssistant Professor, Manav Rachna International Institute of Research and Studies, Sector 43, Faridabad, Haryana 121004, India.SonalPathakAssistant Professor, Manav Rachna International Institute of Research and Studies, Sector 43, Faridabad, Haryana 121004, India.ManikRakhraAssistant Professor, School of Computer Science and Engineering, Lovely Professional University, Phagwara, India.0000-0003-1680-6992Kunchanapaalli RamaKrishnaProfessor, Department of Computer Science and Information Technology, K L Deemed to be University, Vaddeswaram 522502, Andhra Pradesh, India.Journal ArticleStays in ICUs are expensive to accommodate and difficult to forecast, yet even a simple forecast can be helpful, as it enables ICUs to allocate beds and staff and schedule transfer times. We address a particular problem of this kind: how much of the length of stay can be determined based on the first day of hospitalization? Our dataset is the publicly available MIMIC-IV Clinical Database Demo, from which we use 140 hospitalization records, retain the 117 that lasted at least one day, and derive 47 features from the first 24 hours of data on vital signs, medication and fluid intake, urine volume, and bedside procedures. Both formulations are addressed for the same set of patients. Exact forecasting of duration turned out to be rather complicated in our case: despite selecting the best-performing regression technique (random forest), the resulting mean absolute error was 2.85 days, barely outperforming the mean baseline predictor and producing a negative value for the coefficient of determination due to a few very long stays contributing disproportionately to the squared error. Yet, what has been accomplished is the ranking, where the Spearman rank correlation.https://jitm.ut.ac.ir/article_108919_76cccc64cad9cdbd33256bf361668073.pdf