Multicasting Strategies for Increasing Network Efficiency in 5G Using Deep Learning

Multicasting in Internet of Things (IoT) includes transfer data from one source to multiple destinations instantaneously. One major issue is the lack of standardized procedures leading to interoperability problems and potential weaknesses in diverse IoT systems. To consider above mentioned limitation this investigate discovers the application of Artificial Intelligence (AI) methods to revolutionize numerous facets of multicast management. Initially, adaptive multicast group management influences real-time data on user mobility and scheme settings through sensors and monitoring tools. Using Deep Q-Networks (DQN) accomplished with Self Organizing Map and Particle Swarm Optimization (SOM-PSO). Secondly, AI-driven resource allocation hires Deep Reinforcement Learning (DRL) to examine traffic patterns and current network loads unceasingly. Third, predictive analytics for multicast traffic demand participates historical data and contextual information by means of the Dynamic Threshold Algorithm with Multi-Link Communication (DTA-MLC). Enhanced edge caching strategies apply Context-aware Long Short-Term Memory models with Graph Neural Networks (C-ALSTM-GNN) to forecast content demand at network edges. Finally, AI-based multicast routing procedures develop Efficient Quality of Service (QoS) Multicast (EQM) trees to enhance routing paths founded on real-time network topology and traffic conditions. The recommended work is implemented by means of network simulator 3.26, and the efficiency of the proposed model is addressed utilizing several performance metrics such as latency, energy efficiency, throughput, packet delivery ratio, traffic prediction rate. The proposed method achieves latency with 32 ms, energy efficiency with 93%, Traffic Prediction rate 96%, throughput with 342 kbps and PDR with 96%