The C K Pithawala College of Engineering and Technology (CKPCET) in Surat, Gujarat, India, is an engineering college, currently affiliated to GTU (Gujarat Technical University), previously a part of the Veer Narmad South Gujarat University system, located on the banks of the Tapti River near Magdalla Port.The college was started in 1998 by the Navyug Trust of Surat which is also home to science, arts, commerce and law colleges of the South Gujarat University (now known as Veer Narmad South Gujarat University). Presently around 1500 students are studying in the institute..
Facing a rise in security threats for new mobile networks, the 3rd-Generation Partnership Project (3GPP) developed 5G-AKA to ensure secure access to 5G services. However, 5G networks are vulnerable to fake base station attacks, where a malicious actor impersonates a legitimate serving network to intercept or tamper with data transmissions between a user and the network. Fake base stations can exploit vulnerabilities in network protocols, intercept sensitive data, disrupt user privacy by tracking their movements, and force connection downgrades to weaker protocols with known vulnerabilities. To combat fake base stations, we introduce a Digital Signature-based Authentication and Key Agreement (DS-AKA) protocol to verify the serving network’s identity. This not only protects against fake base stations but also strengthens overall security. The results from the Automated Validation of Internet Security Protocols and Applications (AVISPA) tool confirm that the proposed protocol is secure and reliable. To ensure the robustness of the proposed authentication protocol, a comprehensive security evaluation was performed. The protocol’s ability to establish a secure session key was formally verified through BAN logic analysis. DS-AKA achieves a better balance between computational overhead and communication costs compared to other existing protocols.
This paper yields analytical and numerical solutions of the one-dimensional Burgers’ equation occurring in longitudinal dispersion phenomenon via porous media. The phenomenon, which can be either miscible or immiscible fluid flow, leads to nonlinear partial differential equations, which are difficult to solve. To overcome this, the work here uses the Homotopy Analysis Method (HAM), a powerful analytical method, to obtain approximate solutions. Moreover, the numerical techniques like Crank–Nicolson Scheme and the B-Spline Collocation Method are used for comparison purposes. Results from all the techniques are in good agreement with each other, and the patterns of convergence are similar. Sufficient boundary conditions are assigned, and graphical plots of the concentration profiles are obtained using Mathematica software (version 12.0). The graphical plots are very effective and accurately represent the reliability and accuracy of the solutions achieved.
This paper is concerned with solving the problem of one-dimensional counter-current imbibition in a homogeneous porous medium. In this research, water and oil are treated as two distinct liquid phases, in which the water is the wetting phase while the oil is the non-wetting phase. This is the common scenario during secondary recovery of oil. During this phase, the fluid behavior is characterized by a nonlinear partial differential equation. To find the solution of this equation, we utilize the Homotopy Analysis Method (HAM), which is a powerful analytical method. Proper boundary conditions are chosen according to the physical phenomenon of the problem. The results are visualized and interpreted with the help of Mathematica 12.0 using graphical plots.
Continual Learning (CL) is crucial in artificial intelligence for systems to maintain relevance and effectiveness by adapting to new data while retaining previously acquired knowledge. This study explores the performance of multiple machine learning algorithms in CL tasks across various stock symbol datasets over different years. The algorithms assessed include decision trees, ridge regression, lasso regression, elastic net regression, random forests, support vector machines, gradient boosting, and Long Short-Term Memory (LSTM). These models are evaluated on their ability to incrementally gather and maintain knowledge over time, crucial for continual learning. Performance is measured using Mean Squared Error (MSE) and R-squared metrics to assess predictive precision and data conformity. Additionally, the evaluation extends to consider stability, flexibility, and scalability—important factors for models operating in dynamic environments. This comprehensive analysis aims to identify which algorithms best support the objectives of continual learning by effectively integrating new information without compromising the integrity of existing knowledge.
Over the past few decades, adhesive joints have become increasingly popular. Adhesive joint has attracted noteworthy consideration in various industries such as aerospace, marine as a replacement for traditional joining methods including riveting, bolting and welding. In adhesive joint, the role of interface geometry plays a vital role for strength of adhesive joint. Non-flat adherend bonded by adhesive contributes significantly to the increase the strength of adhesive joint. The current comprehensive review study's objective is to evaluate several strategies for improving adhesive joint strength along with decreasing the stress concentrations. These strategies are explored in terms of different geometric design, loading condition and material arrangement. The current review also involved different mechanical and fracture testing experimental methods for the adhesive joint which are used to compute fracture parameters of the adhesive joint and also to understand various advantages and disadvantages of various method. Lastly, the paper is intended to provide some thought-provoking guidelines which can be useful for both research and industry are presented.
