Ghosts in the self-accelerating DGP branch with Gauss-Bonnet effect

TL;DR

This paper investigates whether adding a Gauss-Bonnet term to the DGP brane-world model can eliminate ghost instabilities on the self-accelerating branch, concluding that the ghost problem persists despite the modification.

Contribution

It provides a detailed analysis of the ghost issue in the DGP model with Gauss-Bonnet correction, showing the problem cannot be resolved by this extension.

Findings

Ghost excitations remain unavoidable in the modified model.

Gauss-Bonnet term does not cure the ghost instability.

Analysis conducted for a de Sitter brane in an anti-de Sitter bulk.

Abstract

The Dvali-Gabadadze-Porrati brane-world model provides a possible approach to address the late-time cosmic acceleration. However, it has subsequently been pointed out that a ghost instability will arise on the self-accelerating branch. Here, we carefully investigate whether this ghost problem could be possibly cured by introducing the Gauss-Bonnet term in the five-dimensional bulk action, a natural generalization to the Dvali-Gabadadze-Porrati model. Our analysis is carried out for a background where a de Sitter brane is embedded in an anti-de Sitter bulk. Our result shows that the ghost excitations cannot be avoided even in this modified model.