The invention by Dr Rajiv Senapati, Assistant Professor from the Department of Computer Science and Engineering, and his research scholar Dharani Sabari addresses the critical need for efficient and reliable mobile connectivity by introducing a novel system and software framework for real-time cellular traffic prediction. It leverages a lightweight deep learning architecture, specifically designed with a Spatio-Temporal Attention Module (STAM), Efficient Hybrid Attention (EHA), and Depth wise Separable Convolutions (DSC), to accurately capture both short-term usage spikes and long-term network patterns. Read the abstract to learn more about their research invention.

Brief Abstract:
The invention is a novel system and software framework for real-time cellular traffic prediction. It combines a lightweight deep learning architecture with a Spatio-Temporal Attention Module (STAM), Efficient Hybrid Attention (EHA), and Depthwise Separable Convolutions (DSC) to efficiently capture both short-term variations and long-term traffic patterns. The system processes large-scale network data, identifies critical temporal and spatial dependencies, and generates accurate traffic forecasts to enable dynamic resource allocation, load balancing, and congestion control. By integrating algorithmic intelligence with practical network management functionalities, the invention constitutes a hybrid system that merges software-based predictive modeling with network optimization capabilities, suitable for deployment in next-generation 5G and 6G cellular networks.

Explanation in layperson’s terms:
This invention is a smart software system that helps predict mobile network traffic in real time, so cellular networks can work more smoothly and efficiently. It uses advanced yet lightweight artificial intelligence techniques to understand how network usage changes over time and across different locations, learning both short-term spikes and long-term usage patterns. By analyzing large amounts of network data, the system can accurately forecast congestion before it happens and automatically help the network distribute resources better, balance the load, and avoid slowdowns. Overall, it is a software-driven intelligent solution designed to improve performance, reliability, and user experience of modern and future mobile networks such as 5G and upcoming 6G systems.

Practical Implementation:
The practical implementation of this research enables mobile networks to predict and prevent congestion in real time, resulting in faster data speeds, improved service reliability, and efficient resource utilization. Socially, it enhances digital connectivity, supports smart cities and emergency services, and improves overall user experience while reducing energy consumption.

Future Research Plans:
Future research will focus on integrating real-time IoT and user mobility data to further improve traffic prediction accuracy and network adaptability. Additionally, the framework will be extended to support autonomous network management for emerging 6G and AI-native communication systems.