ADL-SensNet: An Adaptive RNN-LSTM Framework for Context-Aware Energy-Efficient Data Transmission in IoT Sensor Networks

Abstract

The widespread deployment of Internet of Things (IoT) sensor networks necessitates adaptive, energy-efficient data transmission mechanisms capable of responding to dynamic network conditions. This paper addresses this challenge by proposing ADL-SensNet, an adaptive RNN-LSTM-based framework for context-aware, energy-efficient data transmission in IoT sensor networks. ADL-SensNet incorporates intelligent clustering, adaptive cluster-head selection, cluster maintenance, and communication strategy selection driven by real-time network parameters, including traffic load, residual energy, Quality of Service (QoS) requirements, resource availability, and network conditions. By dynamically optimizing communication decisions, the framework reduces communication overhead while enhancing network organization and resource utilization. Experimental results demonstrate that ADL-SensNet achieves 88.7% energy efficiency, 90.8% network stability, 87.1% scalability, 93.4% QoS performance, and 93.38% traffic and demand prediction accuracy. Furthermore, it reduces network congestion by 18.3%, power loss by 12.96%, resource utilization overhead by 15.22%, and operational cost by 22.96%. These results confirm that ADL-SensNet offers a sustainable, resilient, and intelligent solution for next-generation IoT sensor networks.