Machine Learning Based Featureless Signalling

Abstract

Direct-sequence spread-spectrum (DSSS) is commonly used to mitigate the effect of jamming and to operate under an adversary receiver's thermal noise floor in order to avoid signal detection. Unfortunately, the discrete nature and unique distribution of DSSS spreading sequences make it relatively easy to detect the resulting transmitted signals. To overcome this issue, this paper proposes a machine learning based scheme that generates featureless, non-repetitive noise-like spread signals. The proposed scheme provides several benefits over the standard DSSS system including the ability to generate signals with low probabilities of detection/intercept, additional processing gain and also an uncoordinated synchronisation method.