Invariant Stochastic Encoders

TL;DR

This paper extends stochastic vector quantisers to develop invariances, enabling effective jammer nulling by automatically discovering and representing only the large-scale jammer components, thus isolating the underlying signal.

Contribution

It introduces an extension to SVQ theory that allows for invariance development, facilitating jammer nulling without prior jammer knowledge.

Findings

Successfully nulls jammers while preserving signals

Automatically discovers jammer properties during training

Requires minimal prior knowledge of jammers

Abstract

The theory of stochastic vector quantisers (SVQ) has been extended to allow the quantiser to develop invariances, so that only "large" degrees of freedom in the input vector are represented in the code. This has been applied to the problem of encoding data vectors which are a superposition of a "large" jammer and a "small" signal, so that only the jammer is represented in the code. This allows the jammer to be subtracted from the total input vector (i.e. the jammer is nulled), leaving a residual that contains only the underlying signal. The main advantage of this approach to jammer nulling is that little prior knowledge of the jammer is assumed, because these properties are automatically discovered by the SVQ as it is trained on examples of input vectors.