IR Inflation from Multiple Branes

TL;DR

The paper analyzes IR inflation driven by a non-Abelian DBI action for N coincident D3-branes in warped throat geometries. It shows that in the large-N limit the dynamics resemble single-brane DBI inflation but with a non-Abelian fuzzy potential that constrains the parameter space, particularly under WMAP constraints, and derives inflationary observables using a Hamilton-Jacobi approach. Finite-N cases (N=2,3) reveal markedly different non-linear kinetic structures, with the speed of sound becoming highly intricate and even vanishing in the N=3 case, yielding distinct non-Gaussian signatures and spectral indices that depend sensitively on Y and the warp structure. Across AdS-type and mass-gap backgrounds, the study finds that achieving the correct amplitude and spectral tilt often requires a low string scale or specific flux configurations, while larger N tends to enhance perturbations, challenging observational viability. The work highlights rich phenomenology arising from non-Abelian worldvolume dynamics and suggests several avenues for embedding in realistic compactifications and exploring reheating and gauge-sector couplings in this framework.

Abstract

In this paper we examine the IR inflation scenario using the DBI action, where we have $N$ multiple branes located near the tip of a warped geometry. At large $N$ the solutions are similar in form to the more traditional single brane models, however we find that it is difficult to simultaneously satisfy the WMAP bounds on the scalar amplitude and the scalar spectral index. We go on to examine two new solutions where N=2 and N=3 respectively, which both have highly non-linear actions. The sound speed in both cases is dramatically different from previous works, and for the N=3 case it can actually be zero. We show that inflation is possible in both frameworks, and find that the scalar spectral index is bounded from above by unity. The level of non-gaussian fluctuations are smaller in the N=2 case compared to the single brane models, whilst those in the N=3 case are much larger.