[Paper Review] Performance of Lattice Coset Codes on a USRP Testbed.

Jinlong Lu, J. Harshan+1 more|arXiv (Cornell University)|Jul 25, 2016

Wireless Communication Security Techniques6 references3 citations

TL;DR

This paper evaluates lattice coset coding for physical-layer security using a USRP-based wiretap channel experiment with three nodes: Alice (transmitter), Bob (receiver), and Eve (eavesdropper). By varying Eve’s position and adjusting transmit randomness, the study demonstrates that coset coding significantly reduces information leakage, with confidentiality improving as Eve moves farther from the intended receiver.

ABSTRACT

We consider an experimental setup of three USRPs that implement a wiretap channel, two USRPs are the legitimate players Alice and Bob, while the third USRP is the eavesdropper, whose position we vary to evaluate information leakage. The experimented channels are close to slow fading channels, and coset coding of lattice constellations is used for transmission, allowing to introduce controlled randomness at the transmitter. Simulation and measurement results show to which extent coset coding can provide confidentiality, as a function of Eve's position, and the amount of randomness used.

Motivation & Objective

• To evaluate the performance of lattice coset coding in providing confidentiality over wireless channels in a real-world testbed.
• To investigate how the eavesdropper’s spatial position affects information leakage in a wiretap channel setup.
• To quantify the impact of transmit randomness on secrecy rates and security guarantees.
• To bridge the gap between theoretical coset coding and practical implementation using software-defined radio (SDR) platforms.

Proposed method

• Three USRPs are deployed to emulate a wiretap channel: two for Alice and Bob (legitimate parties), one for Eve (eavesdropper).
• Lattice constellations are used for signal modulation, with coset coding introducing controlled randomness at the transmitter.
• The eavesdropper’s position is systematically varied to measure changes in information leakage.
• Experiments are conducted over channels approximating slow fading, with both measurement and simulation results analyzed.
• Information leakage is quantified by comparing mutual information at Bob and Eve, using channel state information and signal-to-noise ratio metrics.
• The secrecy rate is derived from the difference in mutual information between Bob and Eve, serving as the primary performance metric.

Experimental results

Research questions

• RQ1 How does the eavesdropper’s distance from the legitimate receiver affect information leakage in a lattice-coded system?
• RQ2 To what extent can transmit randomness in coset coding reduce information leakage to an eavesdropper?
• RQ3 How well do theoretical secrecy performance predictions match real-world measurements on an SDR testbed?
• RQ4 What is the trade-off between spectral efficiency and security in lattice coset coding under practical channel conditions?

Key findings

• Information leakage to the eavesdropper decreases significantly as Eve’s distance from Bob increases.
• Higher levels of transmit randomness in coset coding lead to reduced information leakage, enhancing physical-layer security.
• The experimental results closely align with simulation predictions, validating the practical feasibility of lattice coset coding.
• Coset coding provides measurable secrecy gains even in slow fading channels, confirming its potential for real-world deployment.