This paper explores the ability of the received signal strength indicator (RSSI) versus frequency pattern (the RSSI signature) to reliably quantify the effect of environmental multipath on RSSI-derived distance measurements for passive ultrahigh frequency systems. Radar technology has demonstrated the use of frequency information for range measurements, given an extremely large bandwidth. In this paper, we show the applicability of these concepts to the ultrahigh frequency radio frequency identification spectrum and its narrow bandwidth. First, we present a theoretical model which illustrates the need for the frequency information to separate multipath error from RSSI measurements. Practically, a closed-form method to extract the multipath component using data from a complex environment is not feasible; therefore, a neural network is used to emulate theoretical variable separation to extract measurement error due to multipath via information from the RSSI frequency signature. The subsequent-predicted distance error relationship not only reduces the error magnitude but also informs the direction of the error, thus making it possible to significantly improve the original distance prediction, even in a completely new environment.
|Journal||IEEE Transactions on Instrumentation and Measurement|
|State||Accepted/In press - 2018|
- Electromagnetic reflection
- error compensation
- neural network applications
- radio distance measurement
- radio frequency identification (RFID)
- signal analysis.
ASJC Scopus subject areas
- Electrical and Electronic Engineering