Uncooperative RF Direction Finding with I/Q Data
Vaidyanath Areyur Shanthakumar, Chaity Banerjee, Tathagata Mukherjee, Eduardo L. Pasiliao · 2020
This paper studies the possibility of implicitly exploiting the characteristics of the In-phase and Quadrature components (I/Q components) of a transmitter using deep learning techniques for the problem of uncooperative direction finding using a single un-calibrated directional receiver. Radio "Direction Finding" (DF) is the problem of estimating the direction of a radio transmitter using features of the received signal. In this paper, we study this problem in the 2.4 GHz WiFi band and restrict ourselves to using I/Q information in a deep learning framework. For this work we used a custom designed data acquisition system built with commercial off-the-shelf (COTS) hardware and collected over the air raw I/Q signal data in both indoor and outdoor settings. The experimental results show that it is possible to reliably predict the bearing of the transmitter with an error bounded by 10 degrees in both indoor and outdoor environments. As our goal was to build an end-to-end system for direction finding with the raw I/Q data, we do not explicitly model the multi-path that inevitably arises in such situations and neither do we hand engineer features to mitigate the problems arising out of the same. Since the characteristics of a transmitter's I/Q data does not change in response to changes in the modulation schemes, the proposed approach has the ability to find the direction of specific emitters in-spite of changes to their modulation scheme.