Efficient Training of Deep Classifiers for Wireless Source Identification using Test SNR Estimates

TL;DR

This paper proposes efficient deep learning training methods for wireless source identification that leverage test SNR estimates, reducing training time and improving accuracy across SNR ranges, especially with limited data.

Contribution

It introduces SNR-specific training strategies, including a greedy SNR boosting algorithm and bagging, to enhance deep classifier performance in wireless signal identification.

Findings

Training on test SNR data reduces training time with minimal accuracy loss.

A small positive SNR offset in training improves test accuracy.

SNR boosting and bagging techniques enhance low SNR generalization.

Abstract

We study efficient deep learning training algorithms that process received wireless signals, if a test Signal to Noise Ratio (SNR) estimate is available. We focus on two tasks that facilitate source identification: 1- Identifying the modulation type, 2- Identifying the wireless technology and channel in the 2.4 GHz ISM band. For benchmarking, we rely on recent literature on testing deep learning algorithms against two well-known datasets. We first demonstrate that using training data corresponding only to the test SNR value leads to dramatic reductions in training time while incurring a small loss in average test accuracy, as it improves the accuracy for low SNR values. Further, we show that an erroneous test SNR estimate with a small positive offset is better for training than another having the same error magnitude with a negative offset. Secondly, we introduce a greedy training SNR Boosting algorithm that leads to uniform improvement in accuracy across all tested SNR values, while using a small subset of training SNR values at each test SNR. Finally, we demonstrate the potential of bootstrap aggregating (Bagging) based on training SNR values to improve generalization at low test SNR values with scarcity of training data.