Deep Learning Based G-LSTM Method for Effective Communication in Wireless Sensor Networking

Abstract

Wireless technology is one of the main uses in healthcare sectors. Particularly emerging nations need a inexpensive, dependable network with effective protocols. Heterogeneity is the most difficult issue with Body Area Networks (BANs), since it necessitates dependability and fairness across all network nodes. Proposed solutions for these networks either don't offer equitable packet transport or use a lot of energy and cause delays. In this research, we present a multi-label text classification model called Graph-Long Short-Term Memory (G- LSTM). We utilize a graph database to store the documents in the suggested model. It builds the categorized dictionaries after pre-processing the documents using standard dictionaries. The sub graphs are created using these categorized dictionaries. Because the protocol can be deployed over current wireless infrastructure and offers high network reliability with energy efficiency via collaboration and adaptation, it is also practical for underdeveloped countries. The suggested system outperforms traditional BAN protocols in terms of energy consumption, throughput, reliability, and Packet Delivery Ratio (PDR) for mobile and scalable networks.