A Centralized Infrastructure-Aware Reliable Data Transaction Model in IoT-Enabled MANET and Cloud Using GWO and Attention Mechanism with LSTM

Attention Mechanism with LSTM

Abstract

Despite the advantages of using IoT-enabled Mobile Ad-hoc Networks (MANETs) in smart cities, healthcare, and military operations, transmission needs to be efficient and secure. This investigation introduces a "Centralized Infrastructure-Aware Reliable Data Transaction Model" which calibrates network speed and security in these scenarios. For secure data transmission and low computation cost, Edward prime curve cryptography (EPCC) is used and selects the most efficient cluster heads for managing data routing using Grey Wolf optimization (GWO). It also includes a reactive routing algorithm like Ad-hoc On-demand Distance Vector (AODV) routing which provides the high reliability of data sent and received dynamically by an effective path discover method. It achieves more success rate, efficiency, and privacy than the others, including K-Nearest Neighbor (KNN), Collaborative Computing Trust model CCTM, and Quality of Service QoS in data transfer. The figures show that every element is important (the ablation study achieved a 93% success rate and got the entire model to an accuracy of 92.5%) in this example situation — but it still explains why embeddings are biased by definition! This hybrid approach is ideal for capacity-heavy applications across numerous industries, as it provides a reliable and secure network solution.