CSI-MIMO: An efficient Wi-Fi fingerprinting using Channel State Information with MIMO

The paper presents a novel Wi-Fi fingerprinting system CSI-MIMO that uses a fine-grained information known as Channel State Information (CSI). The proposed CSI-MIMO exploits the frequency diversity and spatial diversity using Multiple Input Multiple Output (MIMO) system. The proposed CSI-MIMO uses either magnitude or complex CSI based on the mobility in an indoor environment. The experimental performance of CSI-MIMO is compared with Fine-grained Fingerprinting system (FIFS), CSI with Single Input Single Output (SISO) and a simple CSI with MIMO. The experimental result shows a significant improvement with an accuracy of 0.98 m in static environment and 0.31 m in dynamic environment with optimal war-driving over existing systems.