RCC Institute of Information Technology (RCCIIT) is a government sponsored engineering college which is located in Kolkata, West Bengal, India. The college was established in 1999. Government aids are given to the institution by Government of West Bengal and is academically affiliated to Maulana Abul Kalam Azad University of Technology. All the faculties of this institution are recruited by the Government of West Bengal..
Agriculture plays a vital role in the life of every people and has the great impact on the economic demand of the Indian civilization. It is being observed that due to the immigration of the farmers, the efficiency of yield production has been degraded and thus affecting the farming process. To overcome certain challenges faced by the farmers, a smart agriculture system has been designed that can regulate and maintain the farming process in an efficient manner. This system utilizes the smart sensors to monitor the crucial factors responsible for the production of yields such as soil moisture, temperature and humidity, and water level. The stored data of the sensors are then being transmitted to the micro-controller that helps in regulating the watering cycles as per predefined thresholds. The paper demonstrates the actual benefits, obstacles, and potentiality of the agricultural system through IoT applications and can be fully monitored either manually or automatically.
The present paper introduces EyeSee, an intelligent opthalmic health monitoring system for comprehensive eye health monitoring system in real-time and all-encompassing eye-care management which leverages embedded sensing, cloud connectivity and machine learning methods. The system combines retinal image analysis with physiological and environmental monitoring sensors to evaluate the critical ophthalmic parameters including cup-to-disc ratio, neuro-retinal rim thickness and ISNT rule compliance. A supervised learning model is used to analyze retinal images and determine the level of glaucoma risk. Results of the analysis are saved in a cloud database and presented in a mobile application that offers real-time monitoring, trend analysis, AI-based analysis and automatic generation of reports. The proposed framework facilitates eye-health assessment that is accessible, low cost and remote, which aids in early diagnosis and prevention of eye diseases.
The present work deals with theoretical investigation on complex interactions between conduction band and spin-orbit split-off valence bands for computing spontaneous emission and optical gain in a double quantum well triple barrier structure, under both transverse electric (TE) and transverse magnetic (TM) modes. The subband structure and transition matrix elements in DQWs are significantly altered by the coupling of electronic wavefunctions across adjacent wells, as reflected through the design wavelength range of 1.55 $\mu\mathrm{m}$ and 0.87 $\mu\mathrm{m}$ respectively. 7000 cm-1 and 10000 cm-1 peak optical gain obtained for TM and TE modes respectively, which is superior compared to published data. Varying dimensional parameters through simulation exhibits the suitability of design parameters for higher optical peak closest to the central wavelength, and that speaks in favor of the proposed study. Simulated findings will help to design optical amplifier, quantum LED and integrated photonic systems.
Parkinson's disease (PD), a progressive neurodegenerative disorder characterized by motor impairments such as tremors, bradykinesia, and rigidity, is clinically challenging to diagnose, particularly in early-stage conditions and resourceconstrained environments. This study explores motor impairments as potential digital biomarkers for Parkinson's disease identification through simple, static drawing tasks, including waves and spirals. We present an automated and scalable approach based on a hybrid deep learning architecture that integrates an XGBoost ensemble classifier with an Xception convolutional neural network (CNN) for robust hierarchical feature extraction. The image dataset undergoes extensive preprocessing and data augmentation to enhance model robustness and simulate real-world variability. The proposed Xception-XGBoost model achieves high classification performance, with strong precision, recall, and F1-scores, and an overall test accuracy of approximately 93%. The model's generalizability on unseen data is validated through comprehensive evaluation metrics, including confusion matrix analysis and log-loss convergence. These findings demonstrate the effectiveness and practical feasibility of the proposed image-based, lightweight screening framework for scalable deployment in telemedicine and remote diagnostic applications.
This paper investigates the performance of an energy-harvesting cooperative spectrum-sharing radio network. The proposed system model includes a pair of primary nodes and a relay that harvests energy from a multi-antenna power beacon, which also operates as a gateway for the relay. The energy beamforming design at the power beacon is based only on statistical channel state information. A realistic nonlinear rectenna model is adopted at the relay to capture both sensitivity and saturation effects, as well as capacitor charging dynamics. Analytical expressions are derived for the outage probabilities of the primary and secondary communication links under Rician fading, and asymptotic analyses are presented for both high- and low-signal-to-noise ratio regimes. The results reveal the presence of outage floors due to energy saturation, as well as reliability trade-offs governed by the energy allocation factor at the relay, nodes distance, and array size. The study provides useful design insights for energy-efficient spectrum-sharing IoT systems envisioned for next generation communication systems.
