Received Signal Strength Quantization for Secure Indoor Positioning via Fingerprinting

The increasingly connected world magnifies the threats to users' location privacy. Encryption protocols offer solutions to privacy concerns, but they are computationally very demanding. A reduction of the bit-length of the Received Signal Strength (RSS) measurements is required for a realistic, privacy-preserving positioning system based on fingerprinting. This paper studies the practical design of quantizers for RSS fingerprinting data and analyses the effect of the quantization on the positioning performance, with several real data sets and positioning algorithms. Our results show that a 4-bit quantization of RSS yields the same positioning accuracy as with unquantized RSS and that 1-bit approaches (i.e. proximity based positioning) are also feasible for certain applications.