Visible to the public INFECT: INconspicuous FEC-based Trojan: A hardware attack on an 802.11a/g wireless network

TitleINFECT: INconspicuous FEC-based Trojan: A hardware attack on an 802.11a/g wireless network
Publication TypeConference Paper
Year of Publication2017
AuthorsSubramani, K. S., Antonopoulos, A., Abotabl, A. A., Nosratinia, A., Makris, Y.
Conference Name2017 IEEE International Symposium on Hardware Oriented Security and Trust (HOST)
Date Publishedmay
ISBN Number978-1-5386-3929-0
Keywords802.11a/g transmitter, 802.11a/g wireless network, Binary phase shift keying, channel coding, channel noise, channel noise profiling, Contamination, Decoding, FEC encoding, forward error correction, hardware attack, inconspicuous FEC-based trojan, INFECT, pubcrawl, radio networks, radio receivers, radio transmitters, Resiliency, Robustness, Scalability, security, Signal to noise ratio, Trojan horses, wireless LAN, wireless openaccess research platform
Abstract

We discuss the threat that hardware Trojans (HTs) impose on wireless networks, along with possible remedies for mitigating the risk. We first present an HT attack on an 802.11a/g transmitter (TX), which exploits Forward Error Correction (FEC) encoding. While FEC seeks to protect the transmitted signal against channel noise, it often offers more protection than needed by the actual channel. This margin is precisely where our HT finds room to stage an attack. We, then, introduce a Trojan-agnostic method which can be applied at the receiver (RX) to detect such attacks. This method monitors the noise distribution, to identify systematic inconsistencies which may be caused by an HT. Lastly, we describe a Wireless open-Access Research Platform (WARP) based experimental setup to investigate the feasibility and effectiveness of the proposed attack and defense. More specifically, we evaluate (i) the ability of a rogue RX to extract the leaked information, while an unsuspecting, legitimate RX accurately recovers the original message and remains oblivious to the attack, and (ii) the ability of channel noise profiling to detect the presence of the HT.

URLhttps://ieeexplore.ieee.org/document/7951804/
DOI10.1109/HST.2017.7951804
Citation Keysubramani_infect:_2017