1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: \documentclass[aps,prd,twocolumn,showpacs,preprintnumbers]{revtex4}
3: %,nofootinbib
4: \usepackage[dvips]{graphicx}
5: \usepackage{bm,latexsym,amsmath,amssymb,amsfonts}
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7: \newcommand*{\cP}{{\cal P}}
8: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
9: \begin{document}
10:
11: \title{
12: Quantum-mechanical generation of gravitational waves in braneworld
13: }
14:
15: \author{Tsutomu~Kobayashi}
16: \email{tsutomu@tap.scphys.kyoto-u.ac.jp}
17: \author{Takahiro~Tanaka}
18: \email{tama@scphys.kyoto-u.ac.jp}
19:
20: \affiliation{
21: Department of Physics, Kyoto University, Kyoto 606-8502, Japan
22: }
23:
24: %\date{\today}
25:
26: \begin{abstract}
27: We study the quantum-mechanical generation of
28: gravitational waves during inflation on a brane
29: embedded in a five-dimensional anti-de Sitter bulk.
30: To make the problem well-posed, we consider the setup in which
31: both initial and final phases are given by a de Sitter brane with
32: different values of the Hubble expansion rate.
33: Assuming that the quantum state is in a de Sitter
34: invariant vacuum in the initial de Sitter phase, we
35: numerically evaluate the amplitude of quantum fluctuations of
36: the growing solution of the zero mode in the final de Sitter phase.
37: We find that
38: the vacuum fluctuations of the initial Kaluza-Klein gravitons
39: as well as of the zero mode gravitons contribute to the
40: final amplitude of the zero mode on small scales,
41: and the power spectrum is quite well approximated
42: by what we call the rescaled spectrum,
43: which is obtained by rescaling the
44: standard four-dimensional calculation following a simple mapping rule.
45: Our results confirm the speculation raised in Ref.~\cite{Kobayashi:2003cn}
46: before.
47:
48:
49: \end{abstract}
50:
51: \pacs{04.50.+h, 11.10.Kk, 98.80.Cq}
52:
53: \preprint{KUNS-1970}
54:
55: \maketitle
56: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
57:
58: \section{Introduction}
59:
60:
61:
62: In recent years
63: there has been a growing attention
64: to braneworld scenarios,
65: in which our observable universe
66: is realized as a brane embedded in
67: a higher dimensional bulk spacetime~\cite{Maartens:2003tw}.
68: Among various models
69: a simple setup called the Randall-Sundrum type II model~\cite{Randall:1999vf}
70: is of particular interest
71: because, despite gravity propagating
72: into an infinite extra dimension,
73: four-dimensional general relativity
74: can be recovered on the brane
75: due to the warped geometry of anti-de Sitter
76: spacetime~\cite{Randall:1999vf, Garriga:1999yh, Shiromizu:1999wj}.
77: Cosmological consequences of the Randall-Sundrum type braneworld
78: have been widely
79: discussed~\cite{FRW_brane, Dark_Radiation, Garriga:1999bq, Hawking:2000kj, bulk_inflaton1, bulk_inflaton2},
80: especially regarding cosmological perturbations~\cite{CP}.
81:
82:
83:
84:
85: The generation and evolution of perturbations
86: are among the most important issues in cosmology
87: because of their direct link to cosmological observations
88: such as the stochastic gravitational wave background
89: and the temperature anisotropy of the cosmic microwave
90: background (CMB),
91: by which we can probe the early universe.
92: While the cosmological perturbation theory
93: in the conventional four-dimensional universe
94: is rather established~\cite{Kodama:1985bj, Mukhanov:1990me},
95: calculating cosmological perturbations in the braneworld
96: still remains to be a difficult problem.
97: Although some attempts have been made
98: concerning scalar
99: perturbations~\cite{Koyama:2002nw, Koyama:2003be, Rhodes:2003ev, Koyama:2004cf, Koyama:2004ap, Yoshiguchi:2004nm},
100: further progress is awaited
101: to give a clear prediction about the CMB anisotropy
102: in the Randall-Sundrum braneworld.
103: Almost the same is true for gravitational wave (tensor)
104: perturbations~\cite{Hawking:2000kj, Langlois:2000ns, Gorbunov:2001ge,
105: Kobayashi:2003cn, Hiramatsu:2003iz, Hiramatsu:2004aa, Ichiki:2003hf,
106: Ichiki:2004sx, Easther:2003re, Battye:2003ks, Battye:2004qw,
107: Tanaka:2004ig, Kobayashi:2004wy}.
108: However, since their generation and evolution depend basically only on
109: the background geometry,
110: they are slightly easier to handle.
111:
112:
113: During inflation super-horizon gravitational wave perturbations
114: are generated from vacuum fluctuations of gravitons.
115: The pure de Sitter braneworld is the special case that
116: allows definite analytical computation of quantum fluctuations.
117: Thanks to the symmetry of the de Sitter group,
118: the perturbation equation becomes separable and
119: can be solved exactly~\cite{Langlois:2000ns}.
120: The time variation of the Hubble parameter generally
121: causes mixing of a massless zero mode and
122: massive Kaluza-Klein modes,
123: and this effect was investigated
124: based on the ``junction'' models,
125: which assume an instantaneous transition
126: from a de Sitter to a Minkowski brane~\cite{Gorbunov:2001ge}
127: or to another de Sitter brane~\cite{Kobayashi:2003cn}.
128: These works discussed the quantum-mechanical generation of
129: gravitational waves within such limited and simplified models.
130:
131: As for the classical evolution of gravitational waves
132: during the radiation dominated epoch, several studies have
133: been done~\cite{Hiramatsu:2003iz, Hiramatsu:2004aa, Ichiki:2003hf,
134: Ichiki:2004sx}, assuming that a given single initial mode for each
135: comoving wave number dominates.
136: The focus of these works is mainly on
137: the evolution of modes which re-enter the horizon
138: in the high-energy regime.
139: While the late time evolution is worked out analytically
140: in Refs.~\cite{Kobayashi:2004wy, Koyama:2004cf}
141: by resorting to low-energy approximation methods.
142:
143: In this paper we consider the generation of primordial gravitational
144: waves
145: during inflation in more general models of the Randall-Sundrum type.
146: For definiteness, we adopt a simple setup in which both initial
147: and final phases are described by de Sitter braneworlds.
148: Two de Sitter phases with different values of the Hubble expansion rate are
149: smoothly interpolated.
150: This work is an extension of the work on the ``junction'' models
151: by the present authors in collaboration with
152: H.~Kudoh~\cite{Kobayashi:2003cn}.
153: In the previous analysis the transition of the Hubble rate was
154: abrupt and the gap was assumed to be infinitesimal.
155: Here we extend the previous results to more general models
156: with smooth transition by
157: using numerical calculations with a refined formulation.
158:
159:
160:
161:
162:
163:
164:
165: This paper is organized as follows.
166: In the next section we briefly summarize past
167: studies~\cite{Langlois:2000ns, Gorbunov:2001ge, Kobayashi:2003cn} on
168: the generation of gravitational waves via quantum fluctuations
169: during inflation in the Randall-Sundrum braneworld,
170: emphasizing the mapping formula introduced in
171: Ref.~\cite{Kobayashi:2003cn}.
172: In Sec.~\ref{formulation} we describe
173: our numerical scheme
174: to investigate the generation of gravitational waves,
175: and then in Sec.~\ref{results} we present results of our calculations.
176: Section~\ref{discussion} is devoted to discussion.
177:
178:
179:
180: %\cite{Langlois:2000ns}
181: %\cite{Gorbunov:2001ge}
182: %\cite{Kobayashi:2003cn}
183:
184: \section{Gravitational waves in inflationary braneworld}\label{past}
185:
186:
187: \subsection{Pure de Sitter brane}
188:
189: The background spacetime that we consider is composed of
190: a five-dimensional AdS bulk, whose metric
191: is given in the Poincar\'{e} coordinates by
192: \begin{eqnarray}
193: ds^2=\frac{\ell^2}{z^2}\left(
194: -dt^2+\delta_{ij}dx^idx^j+dz^2
195: \right),\label{Poincare}
196: \end{eqnarray}
197: and a Friedmann brane at $z=z(t)$.
198: Here $\ell$ is the bulk curvature length, which is
199: constrained by table-top experiments as
200: $\ell \lesssim 0.1$ mm~\cite{Long:2002wn}.
201:
202: First let us consider the generation of gravitational waves from
203: pure de Sitter inflation on the brane~\cite{Langlois:2000ns}.
204: The coordinate system appropriate for the present situation is
205: \begin{eqnarray}
206: ds^2=\frac{\ell^2}{\sinh^2 \xi}\left[
207: \frac{1}{\eta^2}\left(
208: -d\eta^2+\delta_{ij}dx^idx^j
209: \right)
210: +d\xi^2 \right],
211: \end{eqnarray}
212: which is obtained from Eq.~(\ref{Poincare})
213: by a coordinate transformation
214: \begin{eqnarray}
215: &&t=\eta\cosh\xi+t_0,
216: \label{t-dS}\\
217: &&z=-\eta\sinh\xi,
218: \label{z-dS}
219: \end{eqnarray}
220: where $t_0$ is an arbitrary constant and $\eta$ is the conformal time,
221: which is negative.
222: The de Sitter brane is located at
223: \begin{eqnarray}
224: \xi = \xi_b = \mbox{constant},
225: \end{eqnarray}
226: and the Hubble parameter on the brane is given by
227: \begin{eqnarray}
228: H=\ell^{-1}\sinh\xi_b.
229: \end{eqnarray}
230:
231:
232:
233: The gravitational wave perturbations are described by the metric
234: \begin{eqnarray}
235: ds^2=\frac{\ell^2}{\sinh^2\xi}\left\{
236: \frac{1}{\eta^2}\left[
237: -d\eta^2+(\delta_{ij}+h_{ij})dx^idx^j
238: \right]
239: +d\xi^2 \right\},
240: \end{eqnarray}
241: and we decompose the perturbations into the spatial Fourier modes as
242: \begin{eqnarray}
243: h_{ij}=\frac{\sqrt{2}}{(2\pi M_5)^{3/2}}\int d^3k~
244: \phi_{\textbf{k}}(\eta, \xi)
245: e^{i \textbf{k}\cdot\textbf{x}} e_{ij},
246: \end{eqnarray}
247: where $e_{ij}$ is the transverse-traceless polarization tensor
248: and $M_5$ is the fundamental mass scale
249: which is related to the four-dimensional Planck mass $M_{\rm Pl}$
250: by $\ell (M_5)^3=M^2_{{\rm Pl}}$.
251: The prefactor $\sqrt{2}/(M_5)^{3/2}$ is chosen so that
252: the kinetic term for $\phi_{\textbf{k}}$ in the effective action
253: is canonically normalized.
254: Hereafter we will suppress the subscript $\textbf{k}$.
255: The linearized Einstein equations give the following
256: Klein-Gordon-type equation for $\phi$:
257: \begin{eqnarray}
258: \left[\frac{\partial^2}{\partial \eta^2}
259: -\frac{2}{\eta}\frac{\partial}{\partial \eta}+k^2
260: -\frac{\sinh^3\!\xi}{\eta^2}
261: \frac{\partial}{\partial\xi}
262: {1\over \sinh^3\!\xi}\frac{\partial}{\partial\xi}\right]\phi=0,
263: \label{K-G_for_dS}
264: \end{eqnarray}
265: which is obviously separable.
266: Assuming the $Z_2$-symmetry across the brane,
267: the boundary condition on the brane is given by
268: \begin{eqnarray}
269: n^{\mu}\partial_{\mu} \phi|_{{\rm brane}} \propto
270: \partial_{\xi}\phi|_{\xi=\xi_b}=0,
271: \end{eqnarray}
272: where $n^{\mu}$ is an unit normal to the brane.
273: Equation~(\ref{K-G_for_dS}) admits
274: one discrete zero mode,
275: \begin{eqnarray}
276: \phi_0(\eta) = {\cal N}\cdot\ell^{-1/2}\frac{H}{\sqrt{2k}}
277: \left(\eta-\frac{i}{k}\right)e^{-ik\eta},
278: \label{defphi0}
279: \end{eqnarray}
280: where ${\cal N}$ is
281: a normalization constant, which will be fixed later,
282: as well as massive Kaluza-Klein (KK) modes,
283: \begin{eqnarray*}
284: \phi_{\nu}(\eta, \xi) =\psi_{\nu}(\eta)\chi_{\nu}(\xi),
285: \end{eqnarray*}
286: where $\psi_{\nu}(\eta)$ and $\chi_{\nu}(\xi)$ obey
287: \begin{eqnarray}
288: &&\left[
289: \frac{\partial^2}{\partial \eta^2}
290: -\frac{2}{\eta}\frac{\partial}{\partial \eta}+k^2
291: +\frac{\nu^2+9/4}{\eta^2}
292: \right]\psi_{\nu}(\eta)=0,\label{KK-eta}
293: \\
294: &&\left[
295: \sinh^3\!\xi\frac{\partial}{\partial\xi}
296: {1\over \sinh^3\!\xi}\frac{\partial}{\partial\xi} +\nu^2+\frac{9}{4}
297: \right]\chi_{\nu}(\xi)=0,\label{KK-xi}
298: \end{eqnarray}
299: and the corresponding Kaluza-Klein mass is given by
300: \begin{eqnarray}
301: m^2=\left(\nu^2+\frac{9}{4}\right)H^2\geq\frac{9}{4}H^2.
302: \end{eqnarray}
303: Due to the mass gap $\Delta m = 3H/2$~\cite{Garriga:1999bq}
304: between the zero mode and the lowest KK mode,
305: all the massive KK modes decay on the brane at super-horizon scales during inflation.
306: The explicit forms of the KK mode functions are shown in Appendix~\ref{KKmodefn}.
307:
308:
309: In inflationary cosmology, fluctuations in the graviton
310: field (and other fields such as the inflaton) are
311: considered to be generated quantum-mechanically.
312: Following the standard canonical quantization scheme
313: we quantize the graviton field.
314: Treating $\phi$ as an operator, it may be expanded as
315: \begin{eqnarray}
316: \phi = \hat a_{0} \phi_0 + \hat a_{0}^{\dagger}\phi_0^*
317: +\int_0^{\infty}d\nu\left(\hat a_{\nu}\phi_{\nu}
318: + \hat a_{\nu}^{\dagger}\phi_{\nu}^* \right).
319: \end{eqnarray}
320: The annihilation and creation operators $\hat a_{n}$
321: and $\hat a_{n}^{\dagger}$ with $n=0$ or $\nu$
322: satisfy the usual commutation relations
323: and the modes $\phi_0$ and $\phi_{\nu}$ are normalized so that
324: they satisfy
325: \begin{eqnarray}
326: &&(\phi_0 \cdot \phi_0) = -(\phi_0^*\cdot\phi_0^*) = 1,
327: \label{Wronskian_condition}\\
328: &&(\phi_{\nu} \cdot \phi_{\nu'}) = - (\phi_{\nu}^*\cdot\phi_{\nu'}^*)
329: = \delta(\nu-\nu'),
330: \nonumber\\
331: &&(\phi_0 \cdot \phi_{\nu}) = (\phi_{0}^*\cdot\phi_{\nu}^*)
332: =0,
333: \nonumber\\
334: &&(\phi_{n}\cdot\phi_{n'}^*) = 0,
335: \qquad\mbox{for}\quad n,n'=0, \nu,
336: \nonumber
337: \end{eqnarray}
338: where $(~\cdot~)$ is the Wronskian defined
339: by~\cite{Gorbunov:2001ge, Kobayashi:2003cn}
340: \begin{eqnarray}
341: (X\cdot Y):=-2i\int_{\xi_b}^{\infty}d\xi
342: \frac{\ell^3}{\eta^2\sinh^3\xi}\left(
343: X\partial_{\eta}Y^*-Y^*\partial_{\eta}X \right).
344: \nonumber\\
345: \end{eqnarray}
346: From the conditions (\ref{Wronskian_condition}) we can
347: determine the normalization of the zero mode as
348: \begin{eqnarray}
349: {\cal N} = C(\ell H),
350: \end{eqnarray}
351: with
352: \begin{eqnarray}
353: C(x):=\left[
354: \sqrt{1+x^2}+x^2\ln\left(\frac{x}{1+\sqrt{1+x^2}}\right)
355: \right]^{-1/2},
356: \end{eqnarray}
357: which is the same as that introduced in
358: Refs.~\cite{Langlois:2000ns, Gorbunov:2001ge, Kobayashi:2003cn},
359: and behaves like
360: $C^2(x)\approx 1$ for $x\ll 1$ and $C^2(x)\approx 3x/2$ for $x\gg 1$.
361: This function plays an important roll throughout the present paper.
362:
363:
364: The expectation value of the squared amplitude of the vacuum
365: fluctuation in the zero mode is given by
366: \begin{eqnarray}
367: |\phi_0|^2&=&
368: \ell^{-1}C^2(\ell H)\frac{H^2}{2k^3}(1+k^2\eta^2)
369: \nonumber\\
370: &\to&\ell^{-1}C^2(\ell H)\frac{H^2}{2k^3},
371: \end{eqnarray}
372: where the expression in the second line is obtained by
373: evaluating the perturbation in the super-horizon regime/at a late time,
374: and hence this is the amplitude of the \textit{growing mode}.
375: In terms of the power spectrum defined by
376: \begin{eqnarray}
377: \cP := \frac{4\pi k^3}{(2\pi)^3}\cdot\frac{2}{(M_5)^3}|\phi_0|^2,
378: \end{eqnarray}
379: we have
380: \begin{eqnarray}
381: \cP = \frac{2C^2(\ell H)}{M_{\rm Pl}^2}\left(\frac{H}{2\pi}\right)^2,
382: \label{deSitter_5D}
383: \end{eqnarray}
384: where we have used $\ell (M_5)^3=M_{\rm Pl}^2$.
385: This is the standard flat spectrum up to
386: the overall factor $C^2(\ell H)$.
387: In conventional four-dimensional cosmology
388: the power spectrum of the primordial gravitational waves
389: from de Sitter inflation is given by~\cite{Liddle:2000cg}
390: \begin{eqnarray}
391: \cP_{{\rm 4D}} = \frac{2}{M_{\rm Pl}^2}\left(\frac{H}{2\pi}\right)^2.
392: \end{eqnarray}
393: Thus just by rescaling the amplitude in four-dimensional cosmology as
394: \begin{eqnarray}
395: H \mapsto HC(\ell H),
396: \label{mapdS}
397: \end{eqnarray}
398: the braneworld result~(\ref{deSitter_5D}) is exactly obtained.
399:
400:
401:
402: \subsection{The ``junction'' model}
403:
404: Pure de Sitter inflation on the brane described in the previous subsection
405: is a special case where the amplitude of the growing zero mode
406: can be obtained completely analytically.
407: As a next step to understand gravitational waves
408: from more general inflation with $H\neq$ const,
409: a discontinuous change in the Hubble parameter
410: %, joining two de Sitter branes at some transition time $\eta=\eta_0$,
411: was considered in Ref.~\cite{Kobayashi:2003cn}.
412: In such a ``junction'' model the Hubble parameter is given by
413: \begin{eqnarray}
414: H(\eta)=\left\{H_i,\qquad\qquad\qquad \eta<\eta_0,
415: \atop
416: H_f=H_i-\delta H,\quad \eta>\eta_0,\right.\label{dis_Hub}
417: \end{eqnarray}
418: and
419: \begin{eqnarray}
420: \frac{\delta H}{H_i}\ll 1,
421: \end{eqnarray}
422: is assumed.
423:
424: The evolution of gravitational wave perturbations
425: can be analyzed
426: by solving the equation of motion backward
427: using the advanced Green's function~\cite{Gorbunov:2001ge}.
428: A set of mode functions can be constructed
429: for the initial de Sitter stage,
430: and another for the final de Sitter stage,
431: either of which forms a complete orthonormal basis for
432: the graviton wave function.
433: The final (growing) zero mode
434: may be written as a linear combination of
435: the initial zero and KK modes.
436: The Bogoliubov coefficients give the creation rate
437: of final zero mode gravitons from the initial vacuum fluctuations
438: in the Kaluza-Klein modes as well as in the zero mode.
439: Thus the power spectrum
440: may be written as a sum of two separate contributions,
441: \begin{eqnarray}
442: \cP = \cP_0+\cP_{{\rm KK}},
443: \label{sumspectrum}
444: \end{eqnarray}
445: where $\cP_0$ and $\cP_{{\rm KK}}$ are the parts
446: coming from the initial zero and Kaluza-Klein modes, respectively.
447:
448:
449: An important result brought by analyzing
450: this junction model is that $\cP$ is well approximated by
451: the rescaled spectrum $\cP_{{\rm res}}$
452: obtained by using a simple map speculated by
453: the exact result of pure de Sitter inflation [Eq.~(\ref{mapdS})].
454: More precisely,
455: let $\cP_{{\rm 4D}}(k)$ be
456: the gravitational wave spectrum evaluated
457: in the standard four-dimensional inflationary universe
458: with the same time dependence of the Hubble parameter~(\ref{dis_Hub}),
459: and then
460: the rescaled spectrum is defined by
461: \begin{eqnarray}
462: \cP_{{\rm 4D}}=\frac{2}{M_{{\rm Pl}}^2}
463: \left(\frac{h_k}{2\pi}\right)^2
464: \mapsto
465: \cP_{{\rm res}}=\frac{2C^2(\ell h_k)}{M_{{\rm Pl}}^2}
466: \left(\frac{h_k}{2\pi}\right)^2.
467: \label{mapping}
468: \end{eqnarray}
469: Comparing the rescaled spectrum with $\cP$
470: obtained from a five-dimensional calculation, one finds that
471: the difference between these two is suppressed
472: to be second order like~\cite{Kobayashi:2003cn}
473: \begin{eqnarray}
474: \left|
475: \frac{\cP-\cP_{{\rm res}}}{\cP}
476: \right|\lesssim \left(\frac{\delta H}{H_i}\right)^2\ll 1.
477: \label{agreement}
478: \end{eqnarray}
479: It might be worth noting here that
480: the KK contribution $\cP_{{\rm KK}}$ is necessary
481: for realizing this interesting agreement
482: between the braneworld result and the (basically) four-dimensional
483: result especially at high energies $\ell H\gg 1$.
484: The above agreement (\ref{agreement}) raised a speculation that
485: the mapping formula $h_k\mapsto C(\ell h_k)$
486: may work with good accuracy
487: in more general inflation models
488: with a smoothly changing expansion rate.
489: The central purpose of the present paper
490: is to check whether this speculation is correct or not.
491:
492: \vspace{4mm}
493:
494:
495:
496:
497: \section{Formulation}\label{formulation}
498:
499:
500:
501:
502: \subsection{Basic equations}
503:
504:
505:
506:
507: Now let us explain the formulation that we use to
508: study the generation of gravitational waves
509: without assuming a pure de Sitter brane.
510: Our formulation is based on double null coordinates,
511: which are presumably the most convenient
512: for numerical calculations.
513:
514:
515: The metric~(\ref{Poincare})
516: can be rewritten by using double null coordinates
517: \begin{eqnarray}
518: &u=t-z,\\
519: &v=t+z,
520: \end{eqnarray}
521: in the form of
522: \begin{eqnarray}
523: ds^2=\frac{4\ell^2}{(v-u)^2}\left(
524: -dudv+\delta_{ij}dx^idx^j
525: \right).
526: \end{eqnarray}
527: The trajectory of the brane can be
528: specified arbitrarily by
529: \begin{eqnarray}
530: v=q(u).
531: \end{eqnarray}
532: By a further coordinate transformation
533: \begin{eqnarray}
534: &&U=u,\\
535: &&q(V)=v,
536: \end{eqnarray}
537: we obtain
538: \begin{eqnarray}
539: ds^2=\frac{4\ell^2}{[q(V)-U]^2}\left[
540: -q'(V)dUdV+\delta_{ij}dx^idx^j
541: \right],
542: \end{eqnarray}
543: where a prime denotes differentiation with respect to the argument.
544: Now in the new coordinates the position of the brane is simply given by
545: \begin{eqnarray}
546: U=V.
547: \end{eqnarray}
548: We will use this coordinate system for actual numerical calculations.
549:
550:
551: The induced metric on the brane is
552: \begin{eqnarray}
553: ds_b^2=\frac{4\ell^2}{[q(V)-V]^2}\left[
554: -q'(V)dV^2+\delta_{ij}dx^idx^j
555: \right],
556: \end{eqnarray}
557: from which we can read off the conformal time $\eta$ and the scale
558: factor $a$, respectively, as
559: \begin{eqnarray}
560: &&d\eta = \sqrt{q'(V)} ~dV,\label{eta-V}\\
561: &&a = \frac{2\ell}{q(V)-V},\label{a-q}
562: \end{eqnarray}
563: and hence the Hubble parameter on the brane
564: is written as
565: \begin{eqnarray}
566: \ell H = \frac{1}{2\sqrt{q'(V)}}[1-q'(V)],
567: \end{eqnarray}
568: or equivalently
569: \begin{eqnarray}
570: q'(V)=\left(\sqrt{1+\ell^2H^2}-\ell H\right)^2.
571: \label{q-H}
572: \end{eqnarray}
573: Given the Hubble parameter as a function of $\eta$,
574: one can integrate Eqs.~(\ref{eta-V}) and~(\ref{q-H})
575: to obtain $q$ as a function of $V$.
576:
577: When the brane undergoes pure de Sitter inflation
578: (and thus $q'(V)=$ constant),
579: the following relation between $(U, V)$
580: and $(\eta, \xi)$ will be useful:
581: \begin{eqnarray}
582: &&\xi = \xi_b + \frac{1}{2}\ln\left(\frac{t_0-U}{t_0-V} \right),
583: \\
584: &&\eta = -e^{-\xi_b}[(t_0-U)(t_0-V)]^{1/2},
585: \end{eqnarray}
586: and
587: \begin{eqnarray}
588: q(V) = e^{-2\xi_b}(V-t_0)+t_0.
589: \label{qVequal}
590: \end{eqnarray}
591:
592:
593:
594: The Klein-Gordon-type equation for a gravitational wave perturbation $\phi$
595: in the $(U, V)$ coordinates
596: reduces to
597: \begin{eqnarray}
598: \!
599: \left[4\partial_U\partial_V
600: +\frac{6}{q(V)-U}(\partial_V-q'(V)\,\partial_U)
601: +q'(V)\,k^2\right]\phi=0,\!\!\!
602: \cr
603: \label{KG_in_UV}
604: \end{eqnarray}
605: supplemented by the boundary condition
606: \begin{eqnarray}
607: [\partial_U-\partial_V]\phi\bigr\vert_{U=V}=0.
608: \label{bc}
609: \end{eqnarray}
610: The expression for the Wronskian evaluated on a constant $V$
611: hypersurface is given by
612: \begin{eqnarray}
613: (X\cdot Y)={2i}\!\int^{V}_{-\infty}\!\!\!
614: dU\! \left[\frac{2\ell}{q(V)-U}\right]^{3}\!
615: (X\partial_UY^*\!-Y^*\partial_U X),\cr
616: \end{eqnarray}
617: which is independent of the choice of the hypersurface.
618:
619:
620: When inflation on the brane deviates from pure de Sitter one,
621: the decomposition into the zero mode and KK modes becomes
622: rather ambiguous.
623: For this reason
624: we require the initial and final phases of inflation to be pure de Sitter,
625: though arbitrary cosmic expansion is allowed in the intermediate stage.
626: In both de Sitter phases, $q(V)$ can be fit by Eq.~(\ref{qVequal}), and
627: $\xi_b$ and $t_0$ are determined, respectively.
628: Hence we have two sets of de Sitter coordinates $(\eta,\xi)$ and
629: $(\tilde\eta, \tilde\xi)$.
630: We distinguish the coordinates in the final phase by associating
631: them with tilde.
632: In the final de Sitter phase
633: the mode will be well outside the horizon,
634: and hence we expand the graviton field
635: in terms of the growing and decaying zero mode solutions $\tilde\phi_g$
636: and $\tilde\phi_d$ as
637: \begin{eqnarray}
638: \phi =\hat A_g\tilde\phi_g+\hat A_d\tilde\phi_d
639: +
640: \int d\nu\left(\hat A_{\nu}\tilde\phi_{\nu}
641: +\hat A_{\nu}^{\dagger}\tilde\phi_{\nu}^*
642: \right),
643: \end{eqnarray}
644: where
645: \begin{eqnarray}
646: &&\tilde\phi_g := {\rm Im}\left[\tilde \phi_0\right],\\
647: &&\tilde\phi_d := {\rm Re}\left[\tilde \phi_0\right],
648: \end{eqnarray}
649: and the mode functions with tilde are defined in the same
650: way as $\phi_0$ and $\phi_\nu$ with the substitution of
651: $(\tilde\xi,\tilde \eta)$ for $(\xi,\eta)$.
652: It can be easily seen that the growing and the decaying
653: modes are normalized as
654: \begin{eqnarray}
655: \left(\tilde\phi_g\cdot \tilde\phi_d\right)_f=\frac{1}{2},\quad
656: \left(\tilde\phi_g\cdot \tilde\phi_g\right)_f
657: =\left(\tilde\phi_d\cdot \tilde\phi_d\right)_f=0,
658: \label{g-d-Wronski-condi}
659: \end{eqnarray}
660: where subscript $f$ means that the expression is to be evaluated in
661: the final de Sitter phase. Notice that
662: de Sitter mode functions thus defined
663: as functions of $(U, V)$ through $(\xi,\eta)$
664: (or $(\tilde\xi,\tilde\eta)$) do not satisfy the equation of
665: motion outside the initial (or final) de Sitter phase.
666: Back in the initial de Sitter phase the graviton field
667: can be expanded as
668: \begin{eqnarray}
669: \phi=
670: \hat a_{0} \phi_0 + \hat a_{0}^{\dagger}\phi_0^*
671: +\int d\nu\left(\hat a_{\nu}\phi_{\nu}
672: + \hat a_{\nu}^{\dagger}\phi_{\nu}^* \right).
673: \label{initial_mode_decompose}
674: \end{eqnarray}
675: We assume that initially
676: the gravitons are in the de Sitter invariant vacuum state
677: annihilated by $\hat a_0$ and $\hat a_{\nu}$,
678: \begin{eqnarray}
679: \hat a_0|0\rangle = \hat a_{\nu}|0\rangle = 0.
680: \end{eqnarray}
681:
682: We would like to evaluate the expectation value of the
683: squared amplitude of the vacuum fluctuation
684: in the growing mode at a late time,
685: \begin{eqnarray*}
686: \left|\tilde\phi_g\right|^2\langle 0|\hat A_g^2|0\rangle,
687: \end{eqnarray*}
688: or equivalently, the power spectrum,
689: \begin{eqnarray}
690: \cP (k) &=& \frac{4\pi k^3}{(2\pi)^3}\frac{2}{(M_5)^3}\cdot
691: \left|\tilde\phi_g\right|^2
692: \langle 0|\hat A_g^2|0\rangle
693: \nonumber\\
694: &\to &\frac{2C^2(\ell H_f)}{M^2_{{\rm Pl}}} \left( \frac{H_f}{2\pi} \right)^2
695: \langle 0|\hat A_g^2|0\rangle,
696: \end{eqnarray}
697: where $H_f$ is the Hubble parameter in the final de Sitter phase.
698: In order to obtain the final amplitude,
699: it is not necessary to solve
700: the evolution of all the (infinite number of) degrees of freedom with their
701: initial conditions set by Eq.~(\ref{initial_mode_decompose}).
702: In fact, we have only to
703: solve the backward evolution of the final decaying mode,
704: as explained below.
705:
706: Suppose that a solution
707: $\Phi(U ,V)$ is chosen so as to satisfy
708: $\Phi=\tilde\phi_d$ in the final de Sitter phase.
709: From the Wronskian condition~(\ref{g-d-Wronski-condi})
710: we see that
711: \begin{eqnarray*}
712: {1\over 2} \hat A_g
713: &=& \left(\phi\cdot \tilde\phi_d \right)_f
714: = \left(\phi\cdot \Phi\right)\cr
715: &=&(\phi_0\cdot \Phi)_i\,\hat a_0+\int d\nu(\phi_{\nu}\cdot \Phi)_i\,
716: \hat a_{\nu} +\mbox{h.c.}.,
717: \end{eqnarray*}
718: where subscript $i$ means that the expression is to be evaluated in
719: the initial de Sitter phase.
720: In the above we used the fact that the Wronskian is constant in time.
721: Thus we obtain
722: \begin{eqnarray}
723: \langle 0|\hat A_g^2|0\rangle
724: =4\left[
725: |(\phi_0\cdot \Phi)_i|^2+\int d\nu|(\phi_{\nu}\cdot \Phi)_i|^2
726: \right],
727: \end{eqnarray}
728: from which we can calculate the power spectrum
729: of the primordial gravitational waves.
730: It is obvious that the spectrum is written in the form of
731: Eq.~(\ref{sumspectrum}) with
732: \begin{eqnarray}
733: \cP_0&:=& \frac{8 C^2(\ell H_f)}{M^2_{{\rm Pl}}} \left( \frac{H_f}{2\pi} \right)^2
734: |(\phi_0\cdot \Phi)_i|^2,
735: \\
736: \cP_{{\rm KK}}&:=&\frac{8C^2(\ell H_f)}{M^2_{{\rm Pl}}} \left( \frac{H_f}{2\pi} \right)^2
737: \int d\nu|(\phi_{\nu}\cdot \Phi)_i|^2.
738: \end{eqnarray}
739:
740:
741:
742: \subsection{Numerical scheme}
743:
744: %------------------------------------------------------%
745: \begin{figure}[t]
746: \begin{center}
747: \includegraphics[keepaspectratio=true,height=60mm]{scheme.eps}
748: \end{center}
749: \caption{Numerical (backward) evolution scheme.}
750: \label{fig:scheme.eps}
751: \end{figure}
752: %------------------------------------------------------%
753:
754:
755: Here we describe the algorithm that we employ to
756: solve the backward evolution of the gravitational perturbations numerically.
757: We set the boundary conditions for the mode $\Phi$
758: so as to be identical to
759: a decaying zero mode $\tilde\phi_d$ at a time $V=V_f$
760: in the final de Sitter phase.
761: We decompose $\Phi$ as
762: \begin{eqnarray}
763: \Phi(U,V) = \tilde\phi_d(U,V) + \delta\Phi(U,V),
764: \end{eqnarray}
765: and solve the equation for $\delta\Phi(U,V)$ instead of
766: $\Phi(U,V)$.
767: The equation of motion for $\delta\Phi$ is obtained as
768: \begin{eqnarray}
769: &&\!\!\!\!\left[4\partial_U\partial_V
770: +\frac{6}{q(V)-U}(\partial_V-q'(V)\partial_U)
771: +q'(V)k^2\right]\delta\Phi
772: \nonumber\\
773: &&\!\!\!\!=-\left[4\partial_U\partial_V
774: +\frac{6}{q(V)-U}(\partial_V-q'(V)\,\partial_U)
775: +q'(V)\,k^2\right]\tilde\phi_d.
776: \nonumber\\\label{evolution_eq}
777: \end{eqnarray}
778: Since both $\Phi$ and $\tilde\phi_d$ satisfy the boundary condition
779: of the form of Eq.~(\ref{bc}), the boundary condition for $\delta \Phi$
780: is also written as
781: \begin{eqnarray}
782: [\partial_U-\partial_V]\delta\Phi\bigr\vert_{U=V}=0.
783: \label{bc2}
784: \end{eqnarray}
785: We immediately find that we do not have to solve the
786: backward evolution in the final de Sitter phase, since
787: $\delta\Phi$ identically vanishes there.
788:
789: In some cases $\Phi$ is not disturbed so much
790: from its final configuration $\tilde\phi_d(U,V_f)$,
791: especially when we discuss a small deviation from the pure de Sitter
792: case.
793: It is advantageous then to use a small quantity $\delta \Phi$
794: as a variable in numerical calculations.
795: Of course, there is no problem even when
796: $\delta \Phi$ does not stay small.
797:
798: Our scheme to obtain the values of $\delta\Phi$
799: at the initial surface $V=V_i$ is as follows.
800: We give the boundary conditions at $V=V_f$
801: as $f_{1,1}, f_{2,1}, \cdots, f_{N,1} =0$,
802: where
803: \begin{eqnarray*}
804: f_{n,m}:=\delta\Phi(U=V_f-\varepsilon (n-1),V=V_f-\varepsilon (m-1)).
805: \end{eqnarray*}
806: A sketch of the numerical grids
807: is shown in Fig.~\ref{fig:scheme.eps}.
808: Here $N$ is taken to be sufficiently large,
809: and we use the same grid spacing $\varepsilon$
810: both in the $U$ and $V$ directions.
811: At a virtual site $(U,V)=(V_f, V_f-\varepsilon)$ we set
812: \begin{eqnarray}
813: f_{1,2} = f_{2,1},
814: \end{eqnarray}
815: so that the boundary condition~(\ref{bc2})
816: is satisfied at $(V_f-\varepsilon/2, V_f-\varepsilon/2)$.
817: From $f_{1,1}$, $f_{2,1}$, and $f_{1,2}$ we can determine
818: the value of $f_{2,2}$ by using the equation of motion~(\ref{evolution_eq})
819: at $(V_f-\varepsilon/2, V_f-\varepsilon/2)$,
820: and then from $f_{2,1}$, $f_{3,1}$, and $f_{2,2}$ we can determine $f_{3,2}$,
821: and so on.
822: We repeat the same procedure in the subsequent time steps
823: until we obtain $f_{M,M}, f_{M+1,M}, \cdots, f_{N,M}$
824: at $V_i=V_f-\varepsilon(M-1)$.
825:
826:
827:
828:
829:
830:
831:
832: \subsection{A toy model}
833:
834:
835:
836: %------------------------------------------------------%
837: \begin{figure}[t]
838: \begin{center}
839: \includegraphics[keepaspectratio=true,height=76mm]{model.eps}
840: \end{center}
841: \caption{The Hubble parameter and the slow-roll parameter $\epsilon$
842: in our model. This is a plot for the model of Fig.~\ref{fig:high1.eps}.}
843: \label{fig:model.eps}
844: \end{figure}
845: %------------------------------------------------------%
846:
847:
848: As stated above, to make the problem well posed,
849: we consider models
850: which have initial and final de Sitter phases.
851: To perform numerical calculations,
852: as a concrete example, we adopt a simple toy model
853: in which the Hubble parameter is given by
854: an analytic form
855: \begin{eqnarray}
856: H(\eta)=\bar H - \Delta\tanh\left(\frac{\eta-\eta_0}{s}\right),
857: \end{eqnarray}
858: where
859: \begin{eqnarray}
860: \bar H:=\frac{H_i+H_f}{2},\quad
861: \Delta :=\frac{H_i-H_f}{2},
862: \end{eqnarray}
863: with $H_i$ and $H_f$
864: the initial and final values of the Hubble parameter,
865: $s$ is a parameter that controls the smoothness of the transition,
866: and $\eta_0$ indicates a transition time.
867: Taking the initial time $\eta_i$ and the final time $\eta_f$
868: so that $\eta_0-\eta_i\gg s$ and $\eta_f-\eta_0\approx -\eta_0\gg s$,
869: we have $H\approx H_i$ for $\eta \to \eta_i$
870: and $H\approx H_f$ for $\eta \to\eta_f$.
871: The scale factor for this inflation model is given by
872: \begin{eqnarray}
873: a(\eta)=\left\{
874: s\Delta \ln\left[2\cosh\left(\frac{\eta-\eta_0}{s}\right)\right]
875: -\bar H \eta + \Delta \eta_0
876: \right\}^{-1},
877: \nonumber\\
878: \end{eqnarray}
879: where the integration constant is determined by imposing
880: that $a\approx (-H_f\eta)^{-1}$ at $\eta \approx \eta_f \approx 0$.
881: The slow-roll parameter, $\epsilon:=-\partial_{\eta}H/(aH^2)$,
882: is obtained as
883: \begin{eqnarray}
884: \epsilon=
885: \frac{\Delta\{ s\Delta \ln [2\cosh((\eta-\eta_0)/s)]-\bar H\eta+\Delta \eta_0 \}}
886: {s[ \bar H \cosh((\eta-\eta_0)/s)-\Delta \sinh((\eta-\eta_0)/s)]^2},
887: \end{eqnarray}
888: which takes maximum at $\eta=\eta_0$. The maximum value is
889: \begin{eqnarray}
890: \epsilon(\eta_0)=\epsilon_{{\rm max}}
891: =\frac{\Delta}{s\bar H^2a(\eta_0)}.
892: \label{ep_max}
893: \end{eqnarray}
894: The behavior of the Hubble parameter and the slow-roll parameter
895: is shown in Fig.~\ref{fig:model.eps}.
896:
897:
898:
899: Here we should mention the limitation of this simple toy model.
900: For fixed values of $H_i$, $H_f$, and $\epsilon_{{\rm max}}$
901: one can prolong the period of the transition $\sim s$
902: measured in conformal time
903: by taking large $s$.
904: However,
905: using the definition of $\epsilon_{\rm max}$~(\ref{ep_max}),
906: the transition time scale in \textit{proper} time
907: is found to be given by
908: $\sim a(\eta_0)s
909: =\bar H^2\epsilon_{\rm max}/\Delta$.
910: Therefore we cannot change the transition time scale
911: independently of the other parameters,
912: $H_i$, $H_f$, and $\epsilon_{{\rm max}}$.
913:
914:
915: \section{Numerical results}\label{results}
916:
917:
918: %
919: \begin{figure}%[p]
920: \begin{center}
921: \includegraphics[keepaspectratio=true,height=85mm]{low1.eps}
922: \end{center}
923: \caption{(color online).
924: Top panel: power spectra of gravitational waves
925: normalized by $(2C^2(\ell H_i)/M_{{\rm Pl}}^2)(H_i/2\pi)^2$.
926: Red diamonds (upper ones) indicate the result including the
927: contributions from the initial Kaluza-Klein modes,
928: while green diamonds (lower ones)
929: represent the contribution from the initial zero mode.
930: Although the rescaled four-dimensional spectrum
931: is shown by blue diamonds, they are almost hidden by
932: the red ones since the result of the five-dimensional
933: calculation is well approximated by the rescaled four-dimensional spectrum.
934: Bottom panel: difference between the five-dimensional and
935: rescaled four-dimensional spectra.}%
936: \label{fig:low1.eps}
937: \end{figure}
938: %
939: \begin{figure}%[p]
940: \begin{center}
941: \includegraphics[keepaspectratio=true,height=85mm]{low2.eps}
942: \end{center}
943: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
944: \label{fig:low2.eps}
945: \end{figure}
946: %
947: \begin{figure}%[p]
948: \begin{center}
949: \includegraphics[keepaspectratio=true,height=85mm]{middle1.eps}
950: \end{center}
951: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
952: \label{fig:middle1.eps}
953: \end{figure}
954: %
955: \begin{figure}%[p]
956: \begin{center}
957: \includegraphics[keepaspectratio=true,height=85mm]{middle2.eps}
958: \end{center}
959: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
960: \label{fig:middle2.eps}
961: \end{figure}
962: %
963: \begin{figure}%[p]
964: \begin{center}
965: \includegraphics[keepaspectratio=true,height=85mm]{middle3.eps}
966: \end{center}
967: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
968: \label{fig:middle3.eps}
969: \end{figure}
970: %
971: \begin{figure}%[p]
972: \begin{center}
973: \includegraphics[keepaspectratio=true,height=85mm]{middle-high.eps}
974: \end{center}
975: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
976: \label{fig:middle-high.eps}
977: \end{figure}
978: %
979: \begin{figure}%[p]
980: \begin{center}
981: \includegraphics[keepaspectratio=true,height=85mm]{high1.eps}
982: \end{center}
983: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
984: \label{fig:high1.eps}
985: \end{figure}
986: %
987: \begin{figure}%[p]
988: \begin{center}
989: \includegraphics[keepaspectratio=true,height=85mm]{high2.eps}
990: \end{center}
991: \caption{Same as Fig.~\ref{fig:low1.eps}, but the parameters are different.}%
992: \label{fig:high2.eps}
993: \end{figure}
994:
995:
996: %------------------------
997: \begin{figure}
998: \begin{center}
999: \includegraphics[keepaspectratio=true,height=35mm]{KK.eps}
1000: \end{center}
1001: \caption{Wronskian $|(\phi_{\nu}\cdot\Phi)_i|^2$ as a function of $\nu$.
1002: The parameters are given by those of Fig.~\ref{fig:high1.eps} and
1003: $k/a(\eta_0)\bar H=10$.}%
1004: \label{fig:KK.eps}
1005: \end{figure}
1006: %------------------------
1007:
1008: \begin{table}
1009: \caption{\label{tab:table1}
1010: Parameters used for the numerical calculations
1011: presented in the figures.
1012: }
1013: \begin{ruledtabular}
1014: \begin{tabular}{l|rrr|rr}
1015: &$\ell H_i$&$\ell H_f$&$\epsilon_{{\rm max}}$&$\ell/s$&$\eta_0/\ell$\\
1016: \hline
1017: Figure~\ref{fig:low1.eps} & 0.11 & 0.10 & 0.21 & 0.09 & $-50.45$\\
1018: Figure~\ref{fig:low2.eps} & 0.11 & 0.08 & 0.69 & 0.1 & $-50.46$\\
1019: Figure~\ref{fig:middle1.eps} & 1.1 & 1.0 & 0.21 & 0.09 & $-50.21$\\
1020: Figure~\ref{fig:middle2.eps} & 1.25 & 1.0 & 0.45 & 0.09 & $-50.21$\\
1021: Figure~\ref{fig:middle3.eps} & 1.5 & 1.0 & 0.75 & 0.09 & $-50.21$\\
1022: Figure~\ref{fig:middle-high.eps} & 3.5 & 3.0 & 0.34 & 0.095 & $-50.08$\\
1023: Figure~\ref{fig:high1.eps} & 11 & 10 & 0.19 & 0.085 & $-50.02$\\
1024: Figure~\ref{fig:high2.eps} & 15 & 10 & 0.75 & 0.09 & $-50.02$\\
1025: \end{tabular}
1026: \end{ruledtabular}
1027: \end{table}
1028:
1029:
1030:
1031: Using various parameters shown in Table.~\ref{tab:table1}
1032: we performed numerical calculations,
1033: the results of which are presented in
1034: Figs.~\ref{fig:low1.eps}--\ref{fig:high2.eps}.
1035: Numbers of grids are $N\times M=50000 \times 1000$, and
1036: the grid separation $\varepsilon$ is chosen to be
1037: $(0.11 - 2.5)\times \ell$
1038: depending on the energy scale of
1039: inflation.
1040: Note here that the step width in conformal time $\eta$
1041: is given by $\Delta\eta=\sqrt{q'(V)}~\varepsilon$ [Eq.~(\ref{eta-V})],
1042: and $q'(V)$ becomes smaller for a larger value of $\ell H$ [Eq.~(\ref{q-H})].
1043: Our choice of $\varepsilon$ makes $\Delta\eta$
1044: about the same size, $\Delta\eta\approx 0.1$,
1045: in all the calculations.
1046: Integration over $\nu$ is performed up to $\nu=20$ -- $40$
1047: with an equal grid spacing of $0.15$.
1048: The typical behavior of the integrand $|(\phi_{\nu}\cdot \Phi)_i|^2$
1049: is shown in Fig.~\ref{fig:KK.eps}.
1050:
1051:
1052: In order to compare the five-dimensional power spectrum
1053: with a four-dimensional counterpart,
1054: we solve the conventional evolution equation
1055: for gravitational waves in four-dimensions,
1056: \begin{eqnarray}
1057: \left(\partial_{\eta}^2+2aH\partial_{\eta} +k^2\right)\Phi =0,
1058: \end{eqnarray}
1059: where $H(\eta)$ is given by the same function
1060: as used for the corresponding five-dimensional computation,
1061: and then we calculate the \textit{rescaled} power spectrum
1062: obtained from the four-dimensional `bare' spectrum
1063: by using the mapping formula (\ref{mapping}).
1064:
1065:
1066: We introduce a parameter that represents
1067: a difference between the five-dimensional
1068: power spectrum and the rescaled four-dimensional spectrum:
1069: \begin{eqnarray}
1070: \delta(k) := \frac{\cP_{{\rm res}} - \cP}{\cP}.
1071: \end{eqnarray}
1072: In all the cases of our calculations we clearly see that
1073: $|\delta| \ll 1$ and thus we conclude that
1074: \textit{the primordial spectrum of the gravitational waves
1075: in the braneworld
1076: is quite well approximated by the rescaled four-dimensional spectrum}.
1077: We also find that the difference becomes larger for larger values of
1078: the slow-roll parameter $\epsilon$,
1079: but $|\delta|$ is no greater than ${\cal O}(10^{-2})$
1080: even for models with $\epsilon_{{\rm max}}\simeq 0.75$.
1081: Notice that the universe is not inflating any more if $\epsilon$ is
1082: greater than unity.
1083: At high energies $\ell H \gtrsim 1$
1084: it can be seen that dependence of $\delta$ on $\ell H$
1085: is weak, while at low energies like $\ell H \sim 0.1$,
1086: $\delta$ is further suppressed compared to the high energy cases.
1087:
1088: Contributions from the initial KK modes can be significant
1089: for small wavelength modes.
1090: For example, for the modes with $k/a(\eta_0)\bar H \simeq 10$
1091: we see that $\cP_{{\rm KK}}/\cP\sim{\cal O}(10^{-2})$
1092: at low energies,
1093: but the KK contribution can become as large as $\cP_{{\rm KK}}/\cP\sim 0.6$
1094: in our most `violent' model with $\ell H\sim 10$ and $\epsilon_{{\rm max}}\simeq 0.75$.
1095:
1096:
1097:
1098: \section{Discussion}\label{discussion}
1099:
1100: In this paper we have investigated
1101: the generation of gravitational waves during inflation on a brane
1102: and computed the primordial spectrum.
1103: Extending the previous work~\cite{Kobayashi:2003cn},
1104: we have considered a model composed of the initial
1105: and final de Sitter stages, and the transition region
1106: connecting them smoothly.
1107: We have numerically solved the backward evolution
1108: of the final \textit{decaying} mode
1109: to obtain the amplitude of the \textit{growing} zero mode
1110: at a late time
1111: by making use of the Wronskian method,
1112:
1113:
1114:
1115: We found that the power spectrum $\cP$
1116: is well approximated by the rescaled spectrum $\cP_{{\rm res}}$
1117: basically calculated in standard four-dimensional inflationary cosmology.
1118: Here the rescaling formula for the amplitude
1119: is given by a simple map $h_k \mapsto h_kC(\ell h_k)$,
1120: where $C$ is the normalization factor of the zero mode.
1121: Although the difference of the two spectra,
1122: $\delta:=(\cP_{{\rm res}}-\cP)/\cP$,
1123: depends on the energy scale of inflation and the slow-roll parameter,
1124: our numerical analysis clearly shows that
1125: in any case the mapping formula works with quite good accuracy,
1126: yielding $|\delta| \lesssim {\cal O}(10^{-2})$.
1127: This implies that the mapping relation
1128: between the two spectra holds
1129: quite generally
1130: in the Randall-Sundrum braneworld,
1131: which we believe is a useful formula.
1132:
1133:
1134:
1135: We have taken into account the
1136: vacuum fluctuations in initial Kaluza-Klein modes as well as the zero mode,
1137: and found that both of them contribute
1138: to the final amplitude of the zero mode.
1139: The amount of the initial KK contribution
1140: can be large on small scales,
1141: and it increases as the energy scale of inflation $\ell H$ becomes
1142: higher.
1143: This gives rise to a quite interesting picture.
1144: When the expansion rate changes during inflation,
1145: zero mode gravitons escape into the bulk as KK gravitons,
1146: but at the same time bulk gravitons
1147: come onto the brane to compensate for the loss,
1148: and these two effects almost cancel each other.
1149: This seems to happen,
1150: irrespective of the energy scale,
1151: in a wide class of the inflation models
1152: even with a not-so-small slow-roll parameter.
1153: It is suggested that this is not the case in the radiation dominated
1154: decelerating
1155: universe~\cite{Hiramatsu:2003iz, Hiramatsu:2004aa, Ichiki:2003hf, Ichiki:2004sx},
1156: where the decay into KK gravitons reduces the amplitude
1157: of the gravitational waves on the brane.
1158: The junction model of Ref.~\cite{Gorbunov:2001ge}
1159: joining de Sitter and Minkowski branes
1160: also show the suppression of the gravitational wave amplitude
1161: at high energies.
1162:
1163:
1164:
1165: What is the reason for the remarkable agreement
1166: of the braneworld spectrum and the rescaled four-dimensional spectrum?
1167: Extremely long wavelength modes,
1168: which leave the horizon
1169: during the initial de Sitter stage much before
1170: the Hubble parameter changes,
1171: have a squared
1172: amplitude of $(2C^2(\ell H_i)/M^2_{{\rm Pl}})(H_i/2\pi)^2$,
1173: and the amplitude of the perturbations stays constant
1174: during the subsequent stages~\footnote{In the $k^2\to 0$ limit,
1175: $\phi$=const is a growing solution of Eq.~(\ref{KG_in_UV}).}.
1176: The same is true for four-dimensional inflationary cosmology, and so
1177: the mapping formula is applicable to these long wavelength modes.
1178: On the other hand, a mode whose wavelength
1179: is much shorter than the Hubble horizon scale at the transition time
1180: will feel the transition as adiabatic and hence the particle production
1181: is exponentially suppressed,
1182: $
1183: \langle 0|\hat A_g^{\dagger} \hat A_g |0\rangle \approx 1,
1184: $
1185: leading to $\cP\approx (2C^2(\ell H_f)/M^2_{{\rm Pl}})(H_f/2\pi)^2$.
1186: The same argument can be applied to the conventional cosmology.
1187: Thus it is not surprising that the mapping formula works
1188: for such short wavelength modes.
1189: However, at present there seems no simple reason
1190: for the amplitude of the modes with $k\sim {\cal O}(a(\eta_0)\bar H)$ to
1191: coincide with the rescaled one.
1192:
1193:
1194:
1195: \acknowledgments
1196: We thank Shinji Mukohyama for valuable discussions.
1197: %T.K. is supported by the JSPS.
1198: This work is supported in part by JSPS Research Fellowships for
1199: Young Scientists No.~1642,
1200: by Grant-in-Aid for Scientific Research, Nos.
1201: 14047212 and 16740141,
1202: and by that for the 21st Century COE
1203: ``Center for Diversity and Universality in Physics'' at
1204: Kyoto university, both from the Ministry of
1205: Education, Culture, Sports, Science and Technology of Japan.
1206:
1207:
1208:
1209: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1210: \appendix
1211:
1212: \section{Kaluza-Klein mode functions}\label{KKmodefn}
1213:
1214:
1215: In the de Sitter braneworld the Kaluza-Klein mode functions
1216: in the time direction satisfy Eq.~(\ref{KK-eta}) and are given by
1217: \begin{eqnarray}
1218: \psi_{\nu}(\eta) = \frac{\sqrt{\pi}}{2}\ell^{-3/2}
1219: e^{-\pi\nu/2}(-\eta)^{3/2}H^{(1)}_{i\nu}(-k\eta),
1220: \end{eqnarray}
1221: where $H^{(1)}_{i\nu}$ is the Hunkel function.
1222: They are normalized so that
1223: \begin{eqnarray}
1224: -\frac{i\ell^3}{\eta^2}\left(
1225: \psi_{\nu}\partial_{\eta}\psi^*_{\nu}-\psi_{\nu}^*\partial_{\eta}\psi_{\nu}
1226: \right) = 1.
1227: \end{eqnarray}
1228: The mode functions in the extra direction satisfies Eq.~(\ref{KK-xi})
1229: and are given by
1230: \begin{eqnarray}
1231: \chi_{\nu}(\xi) &=& C_1(\sinh\xi)^2
1232: \left[
1233: P^{-2}_{-1/2+i\nu}(\cosh\xi)\right.
1234: \nonumber\\
1235: &&\quad\left.-C_2
1236: Q^{-2}_{-1/2+i\nu}(\cosh\xi) \right],
1237: \end{eqnarray}
1238: where $P^{-2}_{-1/2+i\nu}$ and $Q^{-2}_{-1/2+i\nu}$
1239: are the associated Legendre functions.
1240: The constants $C_1$ and $C_2$ are determined by
1241: the normalization condition
1242: \begin{eqnarray}
1243: 2\int^{\infty}_{\xi_b}\frac{d\xi}{(\sinh\xi)^3}\chi^*_{\nu'}\chi_{\nu}
1244: =\delta(\nu-\nu'),
1245: \end{eqnarray}
1246: and the boundary condition~(\ref{bc}), respectively, as
1247: \begin{eqnarray}
1248: C_1&=&\left[\left| \frac{\Gamma(i\nu)}{\Gamma(5/2+i\nu)}\right|^2\right.
1249: \nonumber\\
1250: &&\hspace{-10mm}
1251: \left.+\left| \frac{\Gamma(-i\nu)}{\Gamma(5/2-i\nu)}
1252: -\pi C_2\frac{\Gamma(i\nu-3/2)}{\Gamma(1+i\nu)}\right|^2\right]^{-1/2},
1253: \\
1254: C_2&=&
1255: \frac{P^{-1}_{-1/2+i\nu}(\cosh\xi_b)}{Q^{-1}_{-1/2+i\nu}(\cosh\xi_b)}.
1256: \end{eqnarray}
1257:
1258:
1259:
1260:
1261:
1262: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1263:
1264: %--------- References ---------%
1265: \begin{thebibliography}{99}
1266:
1267: %\cite{Maartens:2003tw}
1268: \bibitem{Maartens:2003tw}
1269: For a nice review, see, e.g.,
1270: R.~Maartens,
1271: %``Brane-world gravity,''
1272: Living Rev.\ Rel.\ {\bf 7}, 7 (2004)
1273: [arXiv:gr-qc/0312059].
1274: %%CITATION = GR-QC 0312059;%%
1275:
1276: %\cite{Randall:1999vf}
1277: \bibitem{Randall:1999vf}
1278: L.~Randall and R.~Sundrum,
1279: %``An alternative to compactification,''
1280: Phys.\ Rev.\ Lett.\ {\bf 83}, 4690 (1999)
1281: [arXiv:hep-th/9906064].
1282: %%CITATION = HEP-TH 9906064;%%
1283:
1284: %\cite{Garriga:1999yh}
1285: \bibitem{Garriga:1999yh}
1286: J.~Garriga and T.~Tanaka,
1287: %``Gravity in the brane-world,''
1288: Phys.\ Rev.\ Lett.\ {\bf 84}, 2778 (2000)
1289: [arXiv:hep-th/9911055],
1290: %%CITATION = HEP-TH 9911055;%%
1291: S.~B.~Giddings, E.~Katz and L.~Randall,
1292: %``Linearized gravity in brane backgrounds,''
1293: JHEP {\bf 0003}, 023 (2000)
1294: [arXiv:hep-th/0002091].
1295: %%CITATION = HEP-TH 0002091;%%
1296:
1297: %\cite{Shiromizu:1999wj}
1298: \bibitem{Shiromizu:1999wj}
1299: T.~Shiromizu, K.~i.~Maeda and M.~Sasaki,
1300: %``The Einstein equations on the 3-brane world,''
1301: Phys.\ Rev.\ D {\bf 62}, 024012 (2000)
1302: [arXiv:gr-qc/9910076].
1303: %%CITATION = GR-QC 9910076;%%
1304:
1305: %****** Cosmology on the brane
1306: %\cite{FRW_brane}
1307: \bibitem{FRW_brane}
1308: P.~Binetruy, C.~Deffayet and D.~Langlois,
1309: %``Non-conventional cosmology from a brane-universe,''
1310: Nucl.\ Phys.\ B {\bf 565}, 269 (2000)
1311: [arXiv:hep-th/9905012],
1312: %%CITATION = HEP-TH 9905012;%%
1313: P.~Binetruy, C.~Deffayet, U.~Ellwanger and D.~Langlois,
1314: %``Brane cosmological evolution in a bulk with cosmological constant,''
1315: Phys.\ Lett.\ B {\bf 477}, 285 (2000)
1316: [arXiv:hep-th/9910219],
1317: %%CITATION = HEP-TH 9910219;%%
1318: S.~Mukohyama,
1319: %``Brane-world solutions, standard cosmology, and dark radiation,''
1320: Phys.\ Lett.\ B {\bf 473}, 241 (2000)
1321: [arXiv:hep-th/9911165],
1322: %%CITATION = HEP-TH 9911165;%%
1323: D.~Ida,
1324: %``Brane-world cosmology,''
1325: JHEP {\bf 0009}, 014 (2000)
1326: [arXiv:gr-qc/9912002],
1327: %%CITATION = GR-QC 9912002;%%
1328: P.~Kraus,
1329: %``Dynamics of anti-de Sitter domain walls,''
1330: JHEP {\bf 9912}, 011 (1999)
1331: [arXiv:hep-th/9910149].
1332: %%CITATION = HEP-TH 9910149;%%
1333:
1334: %\cite{Dark_Radiation}
1335: \bibitem{Dark_Radiation}
1336: A.~Hebecker and J.~March-Russell,
1337: %``Randall-Sundrum II cosmology, AdS/CFT, and the bulk black hole,''
1338: Nucl.\ Phys.\ B {\bf 608}, 375 (2001)
1339: [arXiv:hep-ph/0103214],
1340: %%CITATION = HEP-PH 0103214;%%
1341: D.~Langlois, L.~Sorbo and M.~Rodriguez-Martinez,
1342: %``Cosmology of a brane radiating gravitons into the extra dimension,''
1343: Phys.\ Rev.\ Lett.\ {\bf 89}, 171301 (2002)
1344: [arXiv:hep-th/0206146],
1345: %%CITATION = HEP-TH 0206146;%%
1346: D.~Langlois and L.~Sorbo,
1347: %``Bulk gravitons from a cosmological brane,''
1348: Phys.\ Rev.\ D {\bf 68}, 084006 (2003)
1349: [arXiv:hep-th/0306281],
1350: %%CITATION = HEP-TH 0306281;%%
1351: E.~Leeper, R.~Maartens and C.~F.~Sopuerta,
1352: %``Dynamics of radiating braneworlds,''
1353: Class.\ Quant.\ Grav.\ {\bf 21}, 1125 (2004)
1354: [arXiv:gr-qc/0309080],
1355: %%CITATION = GR-QC 0309080;%%
1356: K.~Ichiki, M.~Yahiro, T.~Kajino, M.~Orito and G.~J.~Mathews,
1357: %``Observational constraints on dark radiation in brane cosmology,''
1358: Phys.\ Rev.\ D {\bf 66}, 043521 (2002)
1359: [arXiv:astro-ph/0203272].
1360: %%CITATION = ASTRO-PH 0203272;%%
1361:
1362: %\cite{Garriga:1999bq}
1363: \bibitem{Garriga:1999bq}
1364: J.~Garriga and M.~Sasaki,
1365: %``Brane-world creation and black holes,''
1366: Phys.\ Rev.\ D {\bf 62}, 043523 (2000)
1367: [arXiv:hep-th/9912118].
1368: %%CITATION = HEP-TH 9912118;%%
1369:
1370: %\cite{Hawking:2000kj}
1371: \bibitem{Hawking:2000kj}
1372: S.~W.~Hawking, T.~Hertog and H.~S.~Reall,
1373: %``Brane new world,''
1374: Phys.\ Rev.\ D {\bf 62}, 043501 (2000)
1375: [arXiv:hep-th/0003052].
1376: %%CITATION = HEP-TH 0003052;%%
1377:
1378: %\cite{bulk_inflaton1}
1379: \bibitem{bulk_inflaton1}
1380: S.~Kobayashi, K.~Koyama and J.~Soda,
1381: %``Quantum fluctuations of bulk inflaton in inflationary brane world,''
1382: Phys.\ Lett.\ B {\bf 501}, 157 (2001)
1383: [arXiv:hep-th/0009160],
1384: %%CITATION = HEP-TH 0009160;%%
1385: Y.~Himemoto and M.~Sasaki,
1386: %``Brane-world inflation without inflaton on the brane,''
1387: Phys.\ Rev.\ D {\bf 63}, 044015 (2001)
1388: [arXiv:gr-qc/0010035],
1389: %%CITATION = GR-QC 0010035;%%
1390: Y.~Himemoto, T.~Tanaka and M.~Sasaki,
1391: %``A bulk scalar in the braneworld can mimic the 4d inflaton dynamics,''
1392: Phys.\ Rev.\ D {\bf 65}, 104020 (2002)
1393: [arXiv:gr-qc/0112027],
1394: %%CITATION = GR-QC 0112027;%%
1395: T.~Tanaka and Y.~Himemoto,
1396: %``Generation of dark radiation in bulk inflaton model,''
1397: Phys.\ Rev.\ D {\bf 67}, 104007 (2003)
1398: [arXiv:gr-qc/0301010].
1399: %%CITATION = GR-QC 0301010;%%
1400:
1401: %\cite{bulk_inflaton2}
1402: \bibitem{bulk_inflaton2}
1403: K.~Koyama and K.~Takahashi,
1404: %``Primordial fluctuations in bulk inflaton model,''
1405: Phys.\ Rev.\ D {\bf 67}, 103503 (2003)
1406: [arXiv:hep-th/0301165],
1407: %%CITATION = HEP-TH 0301165;%%
1408: K.~Koyama and K.~Takahashi,
1409: %``Exactly solvable model for cosmological perturbations in dilatonic brane
1410: %worlds,''
1411: Phys.\ Rev.\ D {\bf 68}, 103512 (2003)
1412: [arXiv:hep-th/0307073],
1413: %%CITATION = HEP-TH 0307073;%%
1414: T.~Kobayashi and T.~Tanaka,
1415: %``Bulk inflaton shadows of vacuum gravity,''
1416: Phys.\ Rev.\ D {\bf 69}, 064037 (2004)
1417: [arXiv:hep-th/0311197].
1418: %%CITATION = HEP-TH 0311197;%%
1419:
1420:
1421:
1422: %* * * * * * * * * * * * * *
1423:
1424: %****** Perturbation (General)
1425: %\cite{CP}
1426: \bibitem{CP}
1427: R.~Maartens, D.~Wands, B.~A.~Bassett and I.~Heard,
1428: %``Chaotic inflation on the brane,''
1429: Phys.\ Rev.\ D {\bf 62}, 041301 (2000)
1430: [arXiv:hep-ph/9912464],
1431: %%CITATION = HEP-PH 9912464;%%
1432: S.~Mukohyama,
1433: %``Gauge-invariant gravitational perturbations of maximally symmetric
1434: %spacetimes,''
1435: Phys.\ Rev.\ D {\bf 62}, 084015 (2000)
1436: [arXiv:hep-th/0004067],
1437: %%CITATION = HEP-TH 0004067;%%
1438: H.~Kodama, A.~Ishibashi and O.~Seto,
1439: %``Brane world cosmology: Gauge-invariant formalism for perturbation,''
1440: Phys.\ Rev.\ D {\bf 62}, 064022 (2000)
1441: [arXiv:hep-th/0004160],
1442: %%CITATION = HEP-TH 0004160;%%
1443: D.~Langlois,
1444: %``Brane cosmological perturbations,''
1445: Phys.\ Rev.\ D {\bf 62}, 126012 (2000)
1446: [arXiv:hep-th/0005025],
1447: %%CITATION = HEP-TH 0005025;%%
1448: D.~Langlois,
1449: %``Evolution of cosmological perturbations in a brane-universe,''
1450: Phys.\ Rev.\ Lett.\ {\bf 86}, 2212 (2001)
1451: [arXiv:hep-th/0010063],
1452: %%CITATION = HEP-TH 0010063;%%
1453: H.~A.~Bridgman, K.~A.~Malik and D.~Wands,
1454: %``Cosmic vorticity on the brane,''
1455: Phys.\ Rev.\ D {\bf 63}, 084012 (2001)
1456: [arXiv:hep-th/0010133],
1457: %%CITATION = HEP-TH 0010133;%%
1458: D.~Langlois, R.~Maartens, M.~Sasaki and D.~Wands,
1459: %``Large-scale cosmological perturbations on the brane,''
1460: Phys.\ Rev.\ D {\bf 63}, 084009 (2001)
1461: [arXiv:hep-th/0012044],
1462: %%CITATION = HEP-TH 0012044;%%
1463: H.~A.~Bridgman, K.~A.~Malik and D.~Wands,
1464: %``Cosmological perturbations in the bulk and on the brane,''
1465: Phys.\ Rev.\ D {\bf 65}, 043502 (2002)
1466: [arXiv:astro-ph/0107245],
1467: %%CITATION = ASTRO-PH 0107245;%%
1468: K.~Koyama and J.~Soda,
1469: %``Evolution of cosmological perturbations in the brane world,''
1470: Phys.\ Rev.\ D {\bf 62}, 123502 (2000)
1471: [arXiv:hep-th/0005239],
1472: %%CITATION = HEP-TH 0005239;%%
1473: K.~Koyama and J.~Soda,
1474: %``Bulk gravitational field and cosmological perturbations on the brane,''
1475: Phys.\ Rev.\ D {\bf 65}, 023514 (2002)
1476: [arXiv:hep-th/0108003],
1477: %%CITATION = HEP-TH 0108003;%%
1478: B.~Gumjudpai, R.~Maartens and C.~Gordon,
1479: %``Density perturbations in a brane-world universe with dark radiation,''
1480: Class.\ Quant.\ Grav.\ {\bf 20}, 3295 (2003)
1481: [arXiv:gr-qc/0304067].
1482: %%CITATION = GR-QC 0304067;%%
1483: %* * * * * * * * * * * * * *
1484:
1485: %****** Observation
1486: %\bibitem{wmap} {\tt http://map.gsfc.nasa.gov/}
1487: %\bibitem{planck} {\tt http://planck.esa.int/}
1488: %\bibitem{lisa} {\tt http://lisa.jpl.nasa.gov/}
1489: %\bibitem{decigo}
1490: %N.~Seto, S.~Kawamura and T.~Nakamura,
1491: %``Possibility of direct measurement of the acceleration of the universe using
1492: %0.1-Hz band laser interferometer gravitational wave antenna in space,''
1493: %Phys.\ Rev.\ Lett.\ {\bf 87}, 221103 (2001)
1494: %[arXiv:astro-ph/0108011].
1495: %\bibitem{bbo}
1496: %S.~Phinney et al, {\em The Big Bang Observer: Direct detection of
1497: %gravitational waves from the birth of the Universe to the Present}, NASA Mission Concept Study.
1498: %* * * * * * * * * * * * * *
1499:
1500: %\cite{Kodama:1985bj}
1501: \bibitem{Kodama:1985bj}
1502: H.~Kodama and M.~Sasaki,
1503: %``Cosmological Perturbation Theory,''
1504: Prog.\ Theor.\ Phys.\ Suppl.\ {\bf 78}, 1 (1984).
1505: %%CITATION = PTPSA,78,1;%%
1506:
1507: %\cite{Mukhanov:1990me}
1508: \bibitem{Mukhanov:1990me}
1509: V.~F.~Mukhanov, H.~A.~Feldman and R.~H.~Brandenberger,
1510: %``Theory Of Cosmological Perturbations. Part 1. Classical Perturbations. Part
1511: %2. Quantum Theory Of Perturbations. Part 3. Extensions,''
1512: Phys.\ Rept.\ {\bf 215}, 203 (1992).
1513: %%CITATION = PRPLC,215,203;%%
1514:
1515: %****** Scalar perturbation
1516:
1517: %\cite{Koyama:2002nw}
1518: \bibitem{Koyama:2002nw}
1519: K.~Koyama,
1520: %``Radion and large scale anisotropy on the brane,''
1521: Phys.\ Rev.\ D {\bf 66}, 084003 (2002)
1522: [arXiv:gr-qc/0204047].
1523: %%CITATION = GR-QC 0204047;%%
1524:
1525: %\cite{Koyama:2003be}
1526: \bibitem{Koyama:2003be}
1527: K.~Koyama,
1528: %``CMB anisotropies in brane worlds,''
1529: Phys.\ Rev.\ Lett.\ {\bf 91}, 221301 (2003)
1530: [arXiv:astro-ph/0303108].
1531: %%CITATION = ASTRO-PH 0303108;%%
1532:
1533: %\cite{Rhodes:2003ev}
1534: \bibitem{Rhodes:2003ev}
1535: C.~S.~Rhodes, C.~van de Bruck, P.~Brax and A.~C.~Davis,
1536: %``CMB anisotropies in the presence of extra dimensions,''
1537: Phys.\ Rev.\ D {\bf 68}, 083511 (2003)
1538: [arXiv:astro-ph/0306343].
1539: %%CITATION = ASTRO-PH 0306343;%%
1540:
1541: %\cite{Koyama:2004cf}
1542: \bibitem{Koyama:2004cf}
1543: K.~Koyama,
1544: %``Late time behavior of cosmological perturbations in a single brane
1545: %model,''
1546: JCAP {\bf 0409}, 010 (2004)
1547: [arXiv:astro-ph/0407263].
1548: %%CITATION = ASTRO-PH 0407263;%%
1549:
1550: %\cite{Koyama:2004ap}
1551: \bibitem{Koyama:2004ap}
1552: K.~Koyama, D.~Langlois, R.~Maartens and D.~Wands,
1553: %``Scalar perturbations from brane-world inflation,''
1554: JCAP {\bf 0411}, 002 (2004)
1555: [arXiv:hep-th/0408222].
1556: %%CITATION = HEP-TH 0408222;%%
1557:
1558: %\cite{Yoshiguchi:2004nm}
1559: \bibitem{Yoshiguchi:2004nm}
1560: H.~Yoshiguchi and K.~Koyama,
1561: %``Quantization of scalar perturbations in brane-world inflation,''
1562: Phys.\ Rev.\ D {\bf 71}, 043519 (2005)
1563: [arXiv:hep-th/0411056].
1564: %%CITATION = HEP-TH 0411056;%%
1565:
1566: %* * * * * * * * * * * * * *
1567:
1568: %****** Tensor perturbation
1569:
1570: %\cite{Langlois:2000ns}
1571: \bibitem{Langlois:2000ns}
1572: D.~Langlois, R.~Maartens and D.~Wands,
1573: %``Gravitational waves from inflation on the brane,''
1574: Phys.\ Lett.\ B {\bf 489}, 259 (2000)
1575: [arXiv:hep-th/0006007].
1576: %%CITATION = HEP-TH 0006007;%%
1577:
1578: %\cite{Gorbunov:2001ge}
1579: \bibitem{Gorbunov:2001ge}
1580: D.~S.~Gorbunov, V.~A.~Rubakov and S.~M.~Sibiryakov,
1581: %``Gravity waves from inflating brane or mirrors moving in adS(5),''
1582: JHEP {\bf 0110}, 015 (2001)
1583: [arXiv:hep-th/0108017].
1584: %%CITATION = HEP-TH 0108017;%%
1585:
1586: %\cite{Kobayashi:2003cn}
1587: \bibitem{Kobayashi:2003cn}
1588: T.~Kobayashi, H.~Kudoh and T.~Tanaka,
1589: %``Primordial gravitational waves in inflationary braneworld,''
1590: Phys.\ Rev.\ D {\bf 68}, 044025 (2003)
1591: [arXiv:gr-qc/0305006].
1592: %%CITATION = GR-QC 0305006;%%
1593:
1594: %\cite{Hiramatsu:2003iz}
1595: \bibitem{Hiramatsu:2003iz}
1596: T.~Hiramatsu, K.~Koyama and A.~Taruya,
1597: %``Evolution of gravitational waves from inflationary brane-world: Numerical
1598: %study of high-energy effects,''
1599: Phys.\ Lett.\ B {\bf 578}, 269 (2004)
1600: [arXiv:hep-th/0308072].
1601: %%CITATION = HEP-TH 0308072;%%
1602:
1603: %\cite{Hiramatsu:2004aa}
1604: \bibitem{Hiramatsu:2004aa}
1605: T.~Hiramatsu, K.~Koyama and A.~Taruya,
1606: %``Evolution of gravitational waves in the high-energy regime of brane-world
1607: %cosmology,''
1608: Phys.\ Lett.\ B {\bf 609}, 133 (2005)
1609: [arXiv:hep-th/0410247].
1610: %%CITATION = HEP-TH 0410247;%%
1611:
1612: %\cite{Ichiki:2003hf}
1613: \bibitem{Ichiki:2003hf}
1614: K.~Ichiki and K.~Nakamura,
1615: %``Causal structure and gravitational waves in brane world cosmology,''
1616: Phys.\ Rev.\ D {\bf 70}, 064017 (2004)
1617: [arXiv:hep-th/0310282].
1618: %%CITATION = HEP-TH 0310282;%%
1619:
1620: %\cite{Ichiki:2004sx}
1621: \bibitem{Ichiki:2004sx}
1622: K.~Ichiki and K.~Nakamura,
1623: %``Stochastic gravitational wave background in brane world cosmology,''
1624: arXiv:astro-ph/0406606.
1625: %%CITATION = ASTRO-PH 0406606;%%
1626:
1627: %\cite{Easther:2003re}
1628: \bibitem{Easther:2003re}
1629: R.~Easther, D.~Langlois, R.~Maartens and D.~Wands,
1630: %``Evolution of gravitational waves in Randall-Sundrum cosmology,''
1631: JCAP {\bf 0310}, 014 (2003)
1632: [arXiv:hep-th/0308078].
1633: %%CITATION = HEP-TH 0308078;%%
1634:
1635: %\cite{Battye:2003ks}
1636: \bibitem{Battye:2003ks}
1637: R.~A.~Battye, C.~Van de Bruck and A.~Mennim,
1638: %``Cosmological tensor perturbations in the Randall-Sundrum model: Evolution
1639: %in the near-brane limit,''
1640: Phys.\ Rev.\ D {\bf 69}, 064040 (2004)
1641: [arXiv:hep-th/0308134].
1642: %%CITATION = HEP-TH 0308134;%%
1643:
1644: %\cite{Battye:2004qw}
1645: \bibitem{Battye:2004qw}
1646: R.~A.~Battye and A.~Mennim,
1647: %``Multiple-scales analysis of cosmological perturbations in brane-worlds,''
1648: Phys.\ Rev.\ D {\bf 70}, 124008 (2004)
1649: [arXiv:hep-th/0408101].
1650: %%CITATION = HEP-TH 0408101;%%
1651:
1652: %\cite{Tanaka:2004ig}
1653: \bibitem{Tanaka:2004ig}
1654: T.~Tanaka,
1655: %``AdS/CFT correspondence in a Friedmann-Lemaitre-Robertson-Walker brane,''
1656: arXiv:gr-qc/0402068.
1657: %%CITATION = GR-QC 0402068;%%
1658:
1659: %\cite{Kobayashi:2004wy}
1660: \bibitem{Kobayashi:2004wy}
1661: T.~Kobayashi and T.~Tanaka,
1662: %``Leading order corrections to the cosmological evolution of tensor
1663: %perturbations in braneworld,''
1664: JCAP {\bf 0410}, 015 (2004)
1665: [arXiv:gr-qc/0408021].
1666: %%CITATION = GR-QC 0408021;%%
1667:
1668: %* * * * * * * * * * * * * *
1669:
1670: %\cite{Long:2002wn}
1671: \bibitem{Long:2002wn}
1672: J.~C.~Long, H.~W.~Chan, A.~B.~Churnside, E.~A.~Gulbis, M.~C.~M.~Varney and J.~C.~Price,
1673: %``New experimental limits on macroscopic forces below 100-microns,''
1674: arXiv:hep-ph/0210004.
1675: %%CITATION = HEP-PH 0210004;%%
1676:
1677: %\cite{Liddle:2000cg}
1678: \bibitem{Liddle:2000cg}
1679: A.~R.~Liddle and D.~H.~Lyth,
1680: ``Cosmological inflation and large-scale structure,''
1681: Cambridge University Press, 2000.
1682: %\href{http://www.slac.stanford.edu/spires/find/hep/www?irn=4458796}{SPIRES entry}
1683:
1684: \end{thebibliography}
1685: %--------- References ---------%
1686:
1687: \end{document}
1688: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%