Energy Optimization in IoT Multi-Path Channels via Deep Learning on G6 Networks

EasyChair Preprint 14207

Abstract

In this research, optimization methods for energy resource allocation are investigated to improve energy efficiency in 6G-IoT networks. By utilizing energy harvesting techniques from hybrid sources such as wind, water, and solar, efforts have been made to provide sustainable energy for IoT networks. However, uncertainty in variable temporal environments does not guarantee continuous connectivity and sustainable energy resources for all network nodes. This research addresses the problem of power allocation in multi-path channels with the aim of identifying influential parameters and employing an LSTM neural network for data-driven predictions based on historical data. The results indicate that the computational load of the algorithm is very low, and optimal responses are achieved by the fourth iteration. As the number of IoT devices increases, the response time grows, and an increase in the Signal-to-Noise and Interference Ratio (SINR) leads to a decrease in energy efficiency. Additionally, an increase in the number of IoT devices results in higher power consumption and reduced efficiency. The average power for a maximum transmission power of 10 megabits per joule has been determined.

Keyphrases: 6G Mobile Network, LSTM neural network, Python, optimal power allocation

@booklet{EasyChair:14207,
  author    = {Rawnaq Abdulmajed and Pouya Derakshan-Barjoei},
  title     = {Energy Optimization in IoT Multi-Path Channels via Deep Learning on G6 Networks},
  howpublished = {EasyChair Preprint 14207},
  year      = {EasyChair, 2024}}