With the evolution of radio astronomy and related education and training, the demand for scalable, efficient, and remote systems in data acquisition, storage, and analysis has significantly increased. Addressing this need, we developed a web interface for a log-periodic dipole antenna array integral to the SKA Test activities at the Gauribidanur Radio Observatory. Built with Streamlit, PyVISA, and SCPI commands, the system enables remote observations, secure data access, logging, interference masking, transit plotting, and fringe analysis. A successful solar transit observation demonstrates its practical value for both observatory operations and educational radio astronomy facilities.

This work introduces a sector-based methodology for star-galaxy classification using SDSS-DR18 data. By segmenting the sky to align with survey observation patterns and training a dedicated convolutional neural network for each sector, the approach improves efficiency and accuracy for astronomical source classification, especially in time-sensitive observational settings.

This preprint presents a score-based diffusion framework for reconstructing the primordial CMB B-mode angular power spectrum from contaminated observations. Using variance-exploding stochastic differential equations and a learned physics-guided prior, the model iteratively denoises and delenses simulated spectra that include gravitational lensing, foregrounds, and instrumental noise, offering a promising path for future CMB polarization missions.