Theoretical entropy assessment of fingerprint-based Wi-Fi localization accuracy.

Highlights: [•] The entropy is employed as a new metric to evaluate the fingerprint-based Wi-Fi location accuracy. [•] The relations of entropy and accuracy with variable RP densities and standard deviations are given. [•] The reckless increase of the number of APs cannot be an effective way to improve location accuracy. [•] The substantial saving of computation cost by the entropy is achieved over the existing metrics. [ABSTRACT FROM AUTHOR]