An Intelligent Framework for Energy Management in IoT Networks Using Machine Learning

The proposal of an intelligent solution for EM of IoT network outlined in this research work is based on ML approach while maintaining the quality of service high. It is a new structure that uses actual and estimated data over the application of forecast analysis and optimization system. Some of the sensors applied in the experiments were integrated in a large IoT system encompassing various devices and collections were made over three months. Specific criteria for the evaluation have been established as the energy to be saved, the actual accuracy of the prediction and its compliance with the established quality of the service. According to this study, there was a clear demonstration that different improvements were accrued at an extent of up to 30 percent of the energy savings in addition to a high value of the prediction error rate of (RMSE 5. 2%, MAE - 3. 8%). This framework proved even more helpful than the traditional approaches to management, which suggests that it is capable of managing energy and the networks of IoT conversely, adding longevity to them. Future work will be focused on the enhancement of the framework extensibility and on implementing the real-time update of the framework’s components, as well as on demonstrating the practical applicability of the TF-Rec on various implementations in order to verify its effectiveness and reliability.