Performance Comparison Between Fingerprinting-based RSS Indoor Localization Techniques at WLAN Frequencies: Simulation Study

This paper presents a performance comparison between two fingerprinting-based received signal strength (RSS) indoor localization techniques at wireless local-area network (WLAN) frequencies: 2.4 GHz and 5.8 GHz. The investigated algorithms include the comparative RSS (CRSS) and vector algorithms. The study was conducted using Wireless InSite ray-tracer software. The simulation was conducted in a simulated environment on the 3rd floor of the Chesham Building, University of Bradford, UK. Also, we presented an estimator which looks at the correlation between the test point RSS and the reference point RSS. The estimator performance is compared to the root mean square error (RMSE) performance. It was found that the CRSS algorithm suffers from the similarity problem while constructing the radio map, and it also suffers from ambiguity problems during localization. The vector algorithm outperforms CRSS algorithms in both frequencies and does not suffer from similarity or ambiguity problems. The proposed estimator shows better performance at both frequencies.