Practical Relay Attack on Contactless Transactions by Using NFC Mobile Phones

Lishoy Francis, Gerhard P. Hancke, Keith Mayes, Konstantinos Markantonakis

TL;DR

Researchers have demonstrated a practical "relay attack" on contactless transactions using nothing but off-the-shelf NFC-enabled mobile phones. By turning one phone into a proxy-reader and another into a proxy-token, they successfully bypassed application-layer security to perform fraudulent payments and e-passport scans over distances of up to 35 meters.

What Problem Does This Paper Address?

Contactless technology (RFID/NFC) operates on a dangerous implicit assumption: if a reader can communicate with a token (like a credit card or key fob), that token must be in close physical proximity (usually <10cm).

A relay attack (or "wormhole attack") exploits this by placing a "proxy-token" near the legitimate reader and a "proxy-reader" near the victim’s actual token. The two proxies then tunnel the communication over a long-distance channel (like Bluetooth or Wi-Fi). To the reader, it looks like the victim is standing right there.

While these attacks have been theorized for years, they previously required custom-built hardware or bulky setups. This paper addresses the gap between theory and practice by proving that modern, unmodified NFC smartphones are now powerful enough to serve as a generic, stealthy relay platform.

Methods

The authors developed a software-only implementation using two then-current mobile platforms: a Nokia 6131 and a BlackBerry 9900.

  1. Proxy-Reader (Nokia): Configured using the JSR 257 Contactless Communication API to act as a standard reader. It captures ISO 14443-4 Application Protocol Data Units (APDUs) from a legitimate smart card.
  2. Proxy-Token (BlackBerry): This was the technical "missing link." The authors utilized the BlackBerry NFC emulation API to create a "soft-SE" (Software Token Emulation). This allowed the phone to mimic a passive contactless token that a real-world reader would accept.
  3. The Relay Channel: The two phones communicated via Bluetooth using the JSR 82 API (specifically the L2CAP protocol). Bluetooth was chosen for its low setup complexity and relatively low latency compared to cellular data.

The software was designed to be application-independent; it simply "bit-pipes" the data, meaning it doesn't need to know if it's relaying a payment, a transit fare, or an ID check.

Datasets

As this is an experimental security paper rather than a machine learning study, there are no training datasets. Instead, the authors evaluated their system against two real-world "targets":

  • Contactless Payment System: A test setup based on first-generation contactless credit cards using static authentication.
  • e-Passport System: A demonstration system that included reading a JPEG photo from a passport record, which tested the system's ability to handle larger data packets.

Key Insights and Ideas

  • The "Soft-SE" Breakthrough: The ability to emulate a token entirely in software is the "aha moment." It removes the need for an attacker to have a physical "programmable" card or custom FPGA hardware.
  • Encryption is Irrelevant: Because the attack happens at the communication layer, it doesn't matter how strong the encryption or authentication protocols are at the application layer. The attacker just relays the encrypted "challenge" and "response" bits back and forth.
  • The Latency Myth: Many believed that the timing requirements of NFC would make relaying over Bluetooth impossible. However, the researchers found that real-world readers are incredibly "patient."
  • Generous Timeouts: While the NFC standard suggests tight timing, the researchers found that a Point-of-Sale (POS) terminal would wait up to 35 seconds for a response, and an e-passport reader would wait over 5 seconds. This provides a massive window for relaying data even over slow networks.

Key Results

  • Successful Relays: The team successfully completed full transactions for both credit cards and e-passports.
  • Range: The attack worked reliably at 15 meters in non-line-of-sight environments and up to 35 meters in open-plan rooms.
  • Latency Overhead: Bluetooth added approximately 50ms of round-trip latency. While this is significantly slower than a direct card-to-reader connection (which takes ~4ms to ~17ms), it was well within the multi-second timeouts of commercial readers.
  • Stealth: Because the attack uses standard smartphones, it is far less likely to be detected by a merchant or bystander compared to someone wielding custom antennas and wires.

Limitations and Future Work

The authors highlight several hurdles for future defense and attack:

  • Proprietary Protocols: Some legacy systems (like Mifare Classic) use non-standard framing or encryption at the hardware level that current NFC phone controllers cannot easily intercept or emulate.
  • Countermeasures:
    • Distance Bounding: The most robust defense is implementing protocols that measure the "Round Trip Time" (RTT) of bits at the speed of light to ensure the token is physically close. However, this requires hardware changes in both readers and cards.
  • Location-Based Security: The paper suggests using the phone's GPS or Cell-ID as a second factor. If the "token" phone and "reader" phone report locations miles apart, the transaction is flagged.
  • Indoor Accuracy: The authors admit that GPS is often unreliable indoors, where most transactions occur, suggesting that better indoor positioning (like Wi-Fi triangulation) would be needed for a robust defense.

Summary updated

Abstract

Contactless technology is widely used in security sensitive applications, including identification, payment and access-control systems. Near Field Communication (NFC) is a short-range contactless technology allowing mobile devices to act primarily as either a reader or a token. Relay attacks exploit the assumption that a contactless token within communication range is in close proximity, by placing a proxy-token in range of a contactless reader and relaying communication over a greater distance to a proxy-reader communicating with the authentic token. It has been theorised that NFC-enabled mobile phones could be used as a generic relay attack platform without any additional hardware, but this has not been successfully demonstrated in practice. We present the first generic practical implementation of a contactless relay attack by using only NFC-enabled mobile phones, requiring only suitable mobile software applications. This implementation reduces the complexity of relay attacks and therefore has potential security implications for current contactless systems.
BibTeX
@inproceedings{francis2012practical,
  author = {Lishoy Francis and Gerhard P. Hancke and Keith Mayes and Konstantinos Markantonakis},
  title = {{Practical Relay Attack on Contactless Transactions by Using NFC Mobile Phones}},
  booktitle = {RFIDSec Asia},
  year = {2012},
  doi = {10.3233/978-1-61499-143-4-21},
}
Figure