A neuro evolutionary scheme for improved IoT energy efficiency in smart cities.

With the emergence of Internet of Things (IoT) and allied applications for smart cities, sustainability goals have seen a prominent emphasis. This paper focuses on the energy efficiency aspect of such sustainable smart city goals. Although energy efficiency has been studied at different levels of a smart city's Information and Communication Technology (ICT) infrastructure, this paper specially focuses on device level energy minimization strategy by means of modelling the energy consumption while accounting for the Clusterheads (CluH) and duty cycling and thereby using evolutionary algorithms. In this paper, a Genetic Algorithm (GA) and a hybrid Artificial Neural Network based Particle Swarm Optimization (PSO), namely Feed Forward Neural Network based PSO(FFNN-PSO) has been used to solve the energy minimization problem. Simulation experiments carried out for different scenarios with varying configuration demonstrate the efficacy of the hybrid neuro evolutionary scheme. [Display omitted] • IoT-based energy optimization model for smart city scenarios. • Genetic Algorithm (GA) and a hybrid Artificial Neural Network based Particle Swarm Optimization (PSO), namely Feed Forward Neural Network based PSO(FFNN-PSO). • Several output metrics, such as the number of alive nodes, load, residual energy, and cost function, were used to pick the best cluster head nodes in IoT network clusters. • The proposed method enacts an intelligent duty cycling by predicting sleep–wake cycles. [ABSTRACT FROM AUTHOR]