Indoor Localization for the Blind Based on the Fusion of a Metaheuristic Algorithm with a Neural Network Using Energy-Efficient WSN

Blind and visually impaired people experience difficulty positioning themselves in unfamiliar areas; they may become disoriented and need help from another person. Referred to as ‘localization’, this task is one of the significant challenges frequently encountered by blind people in unfamiliar indoor and outdoor environments. This paper aims to design and implement an energy-efficient localization system (EELS) for the blind and visually impaired based on a Zigbee wireless network and an artificial neural network (ANN) used in indoor environments. The ANN algorithm was adopted to improve localization accuracy for the blind and visually impaired. The ANN was fused independently with six metaheuristic algorithms, namely backtracking search algorithm (BSA), crow search algorithm (CSA), gravitational search algorithm (GSA), slime mould algorithm (SMA), particle swarm optimization (PSO) and multi-verse optimizer-ANN (MVO), to determine the optimal neurons and learning rate of ANN. Consequently, the ANN’s performance was optimized and localization errors were improved. In addition, the power consumption of the EELS was minimized based on a proposed energy-efficient localization algorithm (EELA). The experimental results in indoor environments verified the potential of MVO-ANN in improving localization errors, achieving mean errors of 0.55 and 0.65 m in testing and validating phases, respectively. In addition, the EELS achieved a power savings of 63% based on EELA. Moreover, the EELS had better localization accuracy and power savings than other systems reported in the literature.