Scenario-based recommendation approach for wireless communications networks of smart meters

Smart meter (SM) systems have different wireless communications infrastructures. Hence, SMs can have varying wireless communications performance under possible scenarios. In this paper, a scenario-based recommendation approach is proposed for wireless communications networks of SMs to optimize the efficiency of communications performance. Thus, the proposed approach enables the determination of the most suitable wireless communications standard and protocol, such as Fifth Generation New Radio (5 G NR) and LoRa, under different scenarios. The goal is to ensure that the SMs used in the target locations and buildings communicate in the most efficient way. To achieve this goal, machine learning (ML) algorithms are exploited while developing a decision mechanism for communications efficiency optimization with high environmental awareness. Moreover, a new synthetic dataset is generated with a computer simulation. The parameters and features in the dataset are formed on the basis of wireless channel characteristics of different wireless communications standards and protocols under different scenarios. Based on the results of ML and comparison for several case studies, it is seen that a beneficial approach is developed for more efficient operation of SMs systems, a significant component of smart grids. The classification accuracy on the dataset have been increased to up to 79% with the suitable ML algorithms. Additionally, a SIR improvement of up to 1 dB has been achieved through the selection of the most suitable wireless communications network for several use cases under different smart grid scenarios. Furthermore, it is important to determine the wireless communications network that changes adaptively to different environmental conditions and requirements while using SMs.