Swipe2Pair: Secure and Fast In-Band Wireless Device Pairing

Yaqi He, Kai Zeng, Long Jiao, Brian L. Mark, Khaled N. Khasawneh

TL;DR

This paper introduces Swipe2Pair, a novel in-band wireless device pairing protocol that achieves secure and fast mutual authentication using only standard wireless interfaces. By leveraging randomized transmission power and analyzing channel fading characteristics, Swipe2Pair effectively distinguishes legitimate nearby devices from remote attackers, enabling a seamless pairing experience in under a second.

What Problem Does This Paper Address?

Connecting new devices to our networks and to each other, especially in the burgeoning Internet of Things (IoT) ecosystem, requires a secure way to establish trust without pre-existing secrets. This process, known as device pairing, is crucial for bootstrapping secure communication. Many existing solutions rely on "out-of-band" (OOB) channels, which often require extra hardware like sensors or specific user interfaces. This limits their applicability to devices that lack these features. While "in-band" methods, using only existing wireless interfaces, offer broader compatibility, they often suffer from usability issues (e.g., requiring multiple user actions) or security vulnerabilities (e.g., needing multiple antennas). The paper aims to bridge this gap by proposing an in-band pairing solution that is both highly secure and user-friendly, requiring minimal intervention and completing the process rapidly.

Methods

Swipe2Pair employs a four-stage protocol:

  1. Public Key Exchange with Randomized Transmission Power: Devices A and B exchange their public keys using the Elliptic Curve Diffie-Hellman (ECDH) protocol. Crucially, each transmission uses a randomized transmission (Tx) power, and both devices record their own Tx power and the received (Rx) power from the other device. This is performed multiple times during a quick swipe motion (under 1 second).
  2. Interlock Protocol for Encrypted Power Information: To prevent attackers from easily manipulating power data, an interlock protocol is used. This protocol encrypts the recorded Tx/Rx power information using a session key derived from the ECDH exchange, ensuring that parts of the message cannot be decrypted without the full message.
  3. Bidirectional Pathloss Computation: Using the exchanged and encrypted power information, both devices compute the pathloss between them. This calculation accounts for distance and channel fading, modeled using a lognormal shadowing model.
  4. Authentication via Valley Shape and Channel Fading Checks:
    • Valley Shape Check: The pathloss data during the swipe motion is expected to exhibit a "valley shape" – decreasing as devices move closer and then increasing. A robust peak-valley detection algorithm is used to identify this characteristic. This check helps distinguish devices that are physically close and moving in a coordinated manner from those that are not.
  5. Channel Fading Variation Check: This is the more sophisticated security check. The paper hypothesizes that remote attackers, even if they can mimic the valley shape, will experience significantly different channel fading characteristics compared to devices in close proximity. This is because multipath effects and environmental shadowing are more pronounced over longer distances. By analyzing the variation in channel fading, Swipe2Pair can detect sophisticated attackers.

Datasets

Not discussed in the paper. The paper relies on experimental data collected in various environments.

Key Insights and Ideas

  • Leveraging Wireless Channel Physics for Security: The core insight is that the physical properties of wireless signal propagation (specifically, channel fading and pathloss) can be exploited for security. Close-proximity devices experience different channel characteristics than distant ones.
  • Randomized Transmission Power as a Deception Countermeasure: Randomizing Tx power is key. It prevents attackers from easily inferring channel fading information, which they could otherwise use to impersonate legitimate devices.
  • Dual-Layer Authentication: Combining a "valley shape" check (for physical proximity and coordinated motion) with a "channel fading variation" check (for distinguishing close devices from distant attackers) provides a robust defense against various attacker types.
  • Usability Through Motion: The "swipe" gesture is an intuitive and fast user interaction that enables the protocol, making it practical for everyday use.

Key Results

  • Speed: The pairing process completes in less than one second.
  • Security: Swipe2Pair demonstrates high security against general, advanced, and even "supreme" attackers (who can precisely estimate location/motion). The channel fading variation check is particularly effective against sophisticated attackers.
  • Usability: The protocol requires only a simple swipe motion with minimal user intervention.
  • Universality: It works with standard wireless interfaces (Wi-Fi, Bluetooth, Zigbee) without requiring extra sensors or hardware.
  • Experimental Validation: The paper presents extensive experimental results in diverse environments (confined office, lobby, dining center) showing high detection rates for attackers, with effectiveness increasing with attacker distance.

Limitations and Future Work

  • Environmental Dependence: The paper notes that the threshold for the channel fading variation check needs to be adapted to different environments, as factors like furniture and human activity influence channel fading. Future work could explore adaptive thresholding.
  • Imperfect Motion Robustness: While the valley shape detection is shown to be robust to some imperfect motions, extreme deviations or very large distances between devices during the swipe could potentially affect detection.
  • Supreme Attacker Assumptions: The analysis assumes the attacker cannot be as close as the legitimate devices, which is a reasonable assumption for physical detection but could be a point of further exploration.
  • Future Work: The authors suggest further exploration of environmental factors influencing channel fading and potentially extending the protocol to other wireless technologies.

Summary updated

Abstract

Wireless device pairing is a critical security mechanism to bootstrap the secure communication between two devices without a pre-shared secret. It has been widely used in many Internet of Things (IoT) applications, such as smart-home and smart-health. Most existing device pairing mechanisms are based on out-of-band channels, e.g., extra sensors or hardware, to validate the proximity of pairing devices. However, out-of-band channels are not universal across all wireless devices, so such a scheme is limited to certain application scenarios or conditions. On the other hand, in-band channel-based device pairing seeks universal applicability by only relying on wireless interfaces. Existing in-band channel-based pairing schemes either require multiple antennas separated by a good distance on one pairing device, which is not feasible in certain scenarios, or require users to repeat multiple sweeps, which is not optimal in terms of usability. Therefore, an in-band wireless device pairing scheme providing high security while maintaining high usability (simple pairing process and minimal user intervention) is highly desired. In this work, we propose an easy-to-use mutual authentication device pairing scheme, named Swipe2Pair, based on the proximity of pairing devices and randomization of wireless transmission power. We conduct extensive security analysis and collect considerable experimental data under various settings across different environments. Experimental results show that Swipe2Pair achieves high security and usability. It only takes less than one second to complete the pairing process with a simple swipe of one device in front of the other.
BibTeX
@inproceedings{he2024swipe,
  author = {Yaqi He and Kai Zeng and Long Jiao and Brian L. Mark and Khaled N. Khasawneh},
  title = {{Swipe2Pair: Secure and Fast In-Band Wireless Device Pairing}},
  booktitle = {WISEC},
  year = {2024},
  doi = {10.1145/3643833.3656127},
}
Figure