Reconstruction of Multi-user Binary Subspace Chirps

Abstract

We consider codebooks of Complex Grassmannian Lines consisting of Binary Subspace Chirps (BSSCs) in $N=2^m$ dimensions. BSSCs are generalizations of Binary Chirps (BCs), their entries are either fourth-roots of unity, or zero. BSSCs consist of a BC in a non-zero subspace, described by an on-off pattern. Exploring the underlying binary symplectic geometry, we provide a unified framework for BSSC reconstruction---both on-off pattern and BC identification are related to stabilizer states of the underlying Heisenberg-Weyl algebra. In a multi-user random access scenario we show feasibility of reliable reconstruction of multiple simultaneously transmitted BSSCs with low complexity.

To appear in IEEE ISIT 2020