An Improved Hybrid Indoor Positioning Algorithm via QPSO and MLP Signal Weighting

Accurate location or positioning of people and self-driven devices in large indoor environments has become an important necessity The application increasingly automated self-operating moving transportation units, spaces demands a precise knowledge their positions. Technologies like WiFi Bluetooth, despite low-cost availability, are sensitive to signal noise fading effects. For these reasons, hybrid approach, which uses two different sources, proven be more resilient accurate for the determination environments. Hence, this paper proposes improved technique implement fingerprinting based positioning, using Received Signal Strength information from available Wireless Local Area Network access points, together with Sensor Networks technology. Six signals were recorded on regular grid anchor covering research space. An optimization was performed by relative weighting, minimize average error over process conducted standard Quantum Particle Swarm Optimization, while position estimate all given sets weighted Multilayer Perceptron (MLP) neural network. Compared our previous works, MLP architecture three hidden layers its learning parameters finely tuned. These experimental results led 20% reduction when suitable set weights calculated process. Our final achieved value 0.725 m incertitude shows sensible improvement compared results.