Implementation of Noise-Shaped Signaling System through Software-Defined Radio

TL;DR

This paper introduces a noise-shaped signaling system implemented via Software-Defined Radio to enhance physical-layer security in wireless communications, demonstrating its effectiveness in low probability of detection and interception scenarios.

Contribution

It proposes a novel ring-shaped noise signaling system, details its SDR implementation, and evaluates its performance using CNN-based metrics, advancing physical-layer security techniques.

Findings

High LPI/LPE communication performance achieved

Eavesdroppers struggle to identify modulation schemes

Performance varies with I/Q data modification settings

Abstract

As developments of electromagnetic weapons, Electronic Warfare (EW) has been rising as the future form of war. Especially in wireless communications, the high security defense systems, such as Low Probability of Detection (LPD), Low Probability of Interception (LPI), or Low Prob-ability of Exploitation (LPE) communication algorithms, are studied to prevent the military force loss. One of the LPD, LPI, and LPE communication algorithm, physical-layer security, has been discussed and studied. We propose a noise signaling system, a type of physical-layer secu-rity, which modifies conventionally modulated I/Q data into a noise-like shape. For presenting the possibility of realistic implementation, we use Software-Defined Radio (SDR). Since there are certain limitations of hardware, we present the limitations, requirements, and preferences of practical implementation of noise signaling system, and the proposed system is ring-shaped signaling. We present the ring-shaped signaling system algorithm, SDR implementation meth-odology, and performance evaluations of the system by the metrics of Bit Error Rate (BER) and Probability of Modulation Identification (PMI), which we obtain by Convolutional Neural Net-work (CNN) algorithm. We conclude that the ring-shaped signaling system can perform a high LPI/LPE communication function due to the eavesdropper cannot obtain the correct used modu-lation scheme information, and the performance can vary by the configurations of the I/Q data modifying factors.