Foundation University Journal of Engineering and Applied Sciences

SDN-based Intrusion Detection and Prevention System Against ARP Spoofing Attacks

2025-10-25T07:22:28+00:00

Software Defined Networking (SDN) separates the control plane from the data plane, enabling centralized configuration through the SDN controller. While this centralization simplifies management, it also makes the controller’s ARP table a critical target, as the stateless nature of ARP allows spoofing attacks. To mitigate this vulnerability, we propose an Intrusion Detection and Prevention System integrated as a controller module. The system monitors ARP and DHCP packets, maintaining a permanent ARP table synchronized with a DHCP table to ensure reliable IP–MAC bindings. The IDPS applies four validation checks in both IP and MAC address scanning modules, ensuring robust detection and prevention of spoofed packets. To achieve scalability, the design employs hashmaps for all lookups, ensuring that each check executes in constant time (O(1)), independent of network size. While this methodology introduces a higher baseline mitigation time (~2.2s) compared to some lightweight approaches, it guarantees predictable performance at scale and comprehensive coverage of spoofing attacks.

Comparative Analysis of GRU and LSTM based Models for Pose Estimation in Pakistan Sign Language Recognition

2025-12-01T12:10:10+00:00

This study explores Sign Language Recognition (SLR) within the context of Pakistan Sign Language (PSL), aiming to bridge communication gaps between signers and non-signers. Sign languages employ handshapes, body gestures, and facial expressions to facilitate communication, addressing the worldwide linguistic needs of deaf communities. While significant efforts have been devoted to global SLR and Sign Language Translation (SLT) systems, limited attention has been paid to PSL. To address this gap, we propose a novel approach for dynamic word-level SLR, incorporating manual and non-manual features. The proposed method utilizes pose estimation RNN-based architectures (GRU and LSTM) on both our proprietary pronoun-based video dataset and the PkSLMNM dataset. By extracting key points from 3D coordinates within individuals, we propose several optimization functions for original and augmented datasets. We then compare the sequential classification potential of GRUs and LSTMs. Our findings reveal that GRU outperforms LSTM, achieving a 4% improvement in real-time classification accuracy on both augmented and original datasets, with an overall accuracy of 98.61%.

A Research-Intensive Framework to Automate the Business Operations of a Smart Water Distribution System

2025-12-16T07:23:03+00:00

Typically, software solutions are developed based on generic assumptions without proper, well-defined research methodologies, leading to applications that may not satisfy the needs of a particular target market. In this paper, a combination of qualitative and quantitative approaches was designed to critically analyze the existing water distribution systems. The research-intensive approach helps to build a framework (Aquarise Intelflow) that caters to the real-world actual issues of the stakeholders. Aquarise Intelflow is a smart technology-based water delivery and distribution framework that provides a rich set of features to address the inefficiencies of existing market solutions. It gives a smooth experience to clients by offering subscription and delivery plans, request/process orders, and tracking deliveries. Vendors are equipped with an interactive management dashboard. Evaluation results of Aquarise Intelflow highlight its performance, including minimum manual intervention with enhanced customer satisfaction. In a nutshell, the proposed solution bridges the existing operational gaps in water distribution systems.

A Multistage CNN with Branch Concatenation for Classification of Dementia Using MRI Data

2025-09-18T12:25:01+00:00

Alzheimer’s Disease (AD) is the most common type of dementia and is caused by the accumulation of amyloid-beta plaques in the brain. Worldwide cases of dementia are expected to triple by 2050, which underscores the importance of early diagnosis. In our work, we proposed a multibranched CNN with three concatenations among the branches and tested the method on a dataset accessed from Kaggle. We also implement the SMOTE algorithm on the dataset to overcome class imbalance. The proposed CNN achieved 99.64% accuracy and 99.89% F1-Score on test data and outperformed the various existing methods. The proposed architecture is special because of its ability to extract intricate features at finer levels. The research paves the pathway for improved treatment plans and better prognosis of AD.

Reinforcement Learning for Optimizing Bio-Composite Processing Conditions under Local Constraints

2025-12-19T09:34:32+00:00

This study proposes a high-performance, reinforcement learning-based optimization framework for bio-composite processing, leveraging the Soft Actor-Critic (SAC) algorithm within a constrained decision-making context. Conventional optimization methods in composite manufacturing often face limitations in balancing multiple interdependent objectives such as mechanical performance, energy efficiency, constraint adherence, and production throughput. To overcome these limitations, this work integrates a high-fidelity digital twin environment with a constrained Markov Decision Process (CMDP) formulation, enabling the SAC agent to learn optimal control strategies in real time while respecting operational boundaries. The proposed model was benchmarked against Proximal Policy Optimization (PPO) and Genetic Algorithm (GA) across four key metrics: tensile strength, energy consumption, constraint violation rate, and cycle time. SAC demonstrated superior performance with a mean tensile strength of 71.5 MPa, energy usage of 1.12 kWh per cycle, and a cycle time of 290 seconds—all achieved with the lowest constraint violation rate of 1.8%. These improvements were statistically validated through one-way ANOVA and Tukey’s HSD tests. Additionally, a 10-fold cross-validation using Latin Hypercube Sampling confirmed the generalizability of the SAC policy under diverse, unseen environmental conditions. The findings substantiate the viability of SAC as a real-time, constraint-sensitive optimizer for advanced composite processing. Its ability to intelligently navigate multi-objective trade-offs and adapt to process variability makes it a promising solution for decentralized and resource-constrained manufacturing environments. This research advances the integration of intelligent control in sustainable materials engineering and sets the stage for future deployment in real-world industrial applications.