Green energy aware and cluster based communication for future load prediction in IoT

Abstract

Listen

Translate article

Green energy management has become a critical economic solution for efficient energy consumption; however, the available literature is deficient in emphasising the importance of edge intelligence in a controllable Internet of Things (IoT). Green energy aware IoT networks are well-suited for a variety of real-time applications such as smart cities, smart homes, smart grids, and industries. Due to the fact that IoT nodes are energy constrained and function on a small internal battery, the development of energy-efficient solutions becomes important. Simultaneously, it is necessary to forecast energy consumption in energy aware IoT networks in order to meet upcoming load. With this in mind, this research proposes a novel green energy-aware cluster communication and future load prediction technique for IoT networks called GEQCC-FLP. The proposed GEQCC-FLP technique’s objective is to determine the effective set of cluster heads (CHs) and forecast the network’s incoming load. The GEQCC-FLP technique consists of two primary stages: clustering using a satin bowerbird optimizer (SBO) and load prediction using a deep random vector functional link network (DRVFLN). Additionally, the constructs a fitness function from three parameters: energy, distance, and delay. Additionally, the Adam optimizer can be used to optimise the DRVFLN model’s hyperparameters, hence improving the prediction outcomes. A DRVFLN-based Adam optimizer with deep random vector functional link networks is used to forecast the network’s incoming load (DRVFLN). This clustering technique is based on the SBO algorithm, which generates three fitness function parameters: energy input, distance travelled, and delay. To demonstrate the enhanced outcome of the GEQCC-FLP technique, a variety of simulations are run and the results examined from a variety of angles. The GEQCC-FLP methodology outperformed the prior methods by 0.563 and 0.687, respectively, in terms of root mean square error (RMSE) and mean absolute percent error (MAPE). The extensive results and discussion support the conclusion that the GEQCC-FLP strategy is an extremely effective method for regulating the usage of renewable energy in Internet of Things networks. The experimental results indicated that the GEQCC-FLP technique outperformed the other techniques.