Messaouda Laidi, Tayeb Bouziani / Cement Wapno Beton 28(3) 169-185 (2023)

AU Islam, T Serseg, K Benarous, F Ahmmed, SMA Kawsar Journal of Molecular Structure, 135999

A Linani, K Benarous, E Erol, L Bou-Salah, T Serseg, M Yousfi Journal of Biomolecular Structure and Dynamics, 1-16

Efficient and economical renewable energy system adapted to the Algerian geothermal context

This paper presents a dynamic transmission power adjustment technique to enhance energy efficiency and extend the operational lifespan of mobile devices in fog computing environments. By monitoring the locations of surrounding access points and their own locations, mobile devices adapt their transmission power prior to task offloading. The technique employs four transmission power levels: 3.6mW (default) for access points beyond 265m, 2.7mW for distances between 210m and 265m, 1.8mW for distances between 150m and 210m, and 0.9m W for distances less than or equal to 150m. Two key metrics are evaluated: the consumed energy of mobile devices (in Joules) and the remaining operational devices over time. Additionally, the number of offloaded tasks for each transmission power level is analyzed. Simulation results demonstrate the effectiveness of dynamic transmission power adjustment in improving energy efficiency, prolonging device lifespan, and optimizing task offloading in fog computing environments. These findings contribute to the advancement of energy-efficient mobile computing and provide valuable insights for future optimizations.

Edge computing (EC) serves as a promising complement to cloud computing, yet the challenge of task offloading decision-making persists. While task offloading can extend device lifespan, it may introduce delays that surpass acceptable application thresholds. To address this challenge, we propose a machine learning (ML) technique that leverages key factors such as network latency, battery level, and device location to inform task offloading decisions. This novel ML technique, which is exclusively presented in this paper, holds potential for enhancing task offloading decision-making. Furthermore, the paper discusses future work, including testing and validating the proposed technique, as well as conducting an investigation into the impact of varying the number of mobile users within an edgeenabled environment.