Effective Hybrid Deep Learning Model of GAN and LSTM for Clustering and Data Aggregation in Wireless Sensor Networks

Title:Effective Hybrid Deep Learning Model of GAN and LSTM for Clustering and Data Aggregation in Wireless Sensor Networks

Volume: 14 Issue: 2

Author(s): K. Hemalatha*M. Amanullah

Affiliation:

• Saveetha Institute of Medical and Technical Sciences, Information Technology, Velappanchavadi, Chennai, India

Keywords: WSN, GAN-based clustering, cluster head selection, LSTM-based data aggregation, performance analysis, quality of service.

Abstract:

Background: Wireless Sensor Networks (WSNs) have emerged as a crucial technology for various applications, but they face a lot of challenges relevant to limited energy resources, delayed communications, and complex data aggregation. To address these issues, this study proposes novel approaches called GAN-based Clustering and LSTM-based Data Aggregation (GCLD) that aim to enhance the performance of WSNs.

Methods:The proposed GCLD method enhances the Quality of Service (QoS) of WSN by leveraging the capabilities of Generative Adversarial Networks (GANs) and the Long Short-Term Memory (LSTM) method. GANs are employed for clustering, where the generator assigns cluster assignments or centroids, and the discriminator distinguishes between real and generated cluster assignments. This adversarial learning process refines the clustering results. Subsequently, LSTM networks are used for data aggregation, capturing temporal dependencies and enabling accurate predictions.

Results: The evaluation results demonstrate the superior performance of GCLD in terms of delay, PDR, energy consumption, and accuracy than the existing methods.

Conclusion: Overall, the significance of GCLD in advancing WSNs highlights its potential impact on various applications.

Cite this article as:

Hemalatha K.*, Amanullah M., Effective Hybrid Deep Learning Model of GAN and LSTM for Clustering and Data Aggregation in Wireless Sensor Networks, International Journal of Sensors, Wireless Communications and Control 2024; 14 (2) . https://dx.doi.org/10.2174/0122103279275330231217072855