Upload
others
View
9
Download
0
Embed Size (px)
Citation preview
Constraints on Tensor to Scalar Ratio using WKB
Approximation
Aiswarya. Aa,b and Minu Joya
a Dept. of Physics, Alphonsa College, Pala 686574, Indiab School of Pure and Applied Physics, Mahatma Gandhi University, Kottayam
686560, India
Abstract. Using Wenzel-Kramers-Brillouin (WKB) approximation the scalar
and tensor power spectra are obtained. Scale invariant spectra are obtained
and the spectral indices come very close to the observed data from WMAP
and Planck experiments. The advantage of this method is that, it is valid even
when slow-roll approximation fails. Constraints on the tensor to scalar ratio(r)
is also studied with the WKB Approximation. We use the Power law inflation
as the base model as it allows comparison with exact results.
arX
iv:1
809.
0857
2v2
[as
tro-
ph.C
O]
18
Feb
2019
Constraints on Tensor to Scalar Ratio using WKB Approximation 2
1. Introduction
Inflationary cosmology has come a long way, starting from being the theory to explain
the shortcomings of Big Bang model [1, 2, 3] to the most accepted model with the current
observations. Success of NASA′s first CMB satellite COBE (Cosmic Background
Explorer satellite), paved way for future projects in CMB. These results gave an accurate
thermal spectrum of CMB, the first detection of primordial CMB fluctuations with
normalization at 10−5. It also gave cosmic variance limit of large angular scale power
spectrum [4]. Among the CMB missions which followed COBE, BICEP mission was
designed to measure the B-mode polarization at degree angular scales and can be
considered as the first mission to constrain the tensor-to-scalar ratio r directly from
CMB B-mode polarization. The mission gave the r constraints as
r = 0.02+0.31−0.26 or r < 0.72 at 95% C.L.
The direct measurement of r from the BB spectrum has advantage, as the
measurements of TT are limited by cosmic variance at large angular scales and the r
constraints from TT have strong parameter degeneracies such as with the scalar spectral
index ns, while the BB amplitude depends primarily on r [5].
The SPT-SZ survey helped to remove the degeneracy between ns and r by
measurements of the power spectrum from angular scales corresponding to multipole
range 650 < l < 3000, that is from the third acoustic peak through the CMB damping
tail. The constrain on scalar spectral index tightens to ns < 1.0 at 3.9σ by using SPT
and WMAP7 data [6]. The B and E mode polarization data gives constraints on the
recombination (last-scattering) and reionization epochs. The cross-correlation between
temperature anisotropies and polarization signal is a consistency check for inflationary
cosmology and helps in breaking degeneracies among the cosmological parameters [7].
Analysis of the combined BICEP2 and Keck Array data in combination with the
2015 Planck data put an upper limit r0.05 < 0.12 at 95% confidence [8]. 2018 release of
the Planck temperature, polarization and lensing data determine the spectral index of
scalar perturbations to be
ns = 0.965± 0.004
at 68% CL and find no evidence for a scale dependence of ns, either as a running or as
a running of the running. The latest Planck 95% CL upper limit on the tensor-to-scalar
ratio, r0.002 < 0.10, is further tightened by combining with the BICEP2/Keck Array
BK14 data to obtain r0.002 < 0.064 [9, 10].
In the present work we study the constraints on the tensor-to-scalar ratio r
using an alternative approximation technique, instead of slow-roll approximation, with
the Power law inflationary model. We first enumerate the impact of r on different
inflationary models and cosmological parameters, followed by a brief review on some
alternative approximation techniques. Scalar and tensor power spectra using the WKB
approximation technique are obtained. Hence we try to constrain tensor to scalar ratio
with WKB approximation and check the consistency with latest Planck observations.
Constraints on Tensor to Scalar Ratio using WKB Approximation 3
2. Tensor to scalar ratio
The study of tensor to scalar ratio r gives insight into large working of the universe :-
• Primordial Gravitational waves - The tensor perturbations in the FLRW
background metric during inflation are called the primordial gravitational waves
and CMB is the only observation to validate it. The CMB sky is measured with
two observables, the temperature anisotropy and the polarisation. Primordial
gravitational waves imprints very slight variations on these variables and are
sensitive to r which is defined as [7]
r =PtPs
It contributes to the temperature anisotropies as a local quadruple since the grav-
itational wave metric is traceless. The primordial tensor power spectrum [11] is
scale invariant and the shear is impulsive at horizon entry [12].
• Energy scale of universe - r depends on the time-evolution of the inflaton field and
the relation to inflaton potential [13] is given as
r =8
M2Pl
( φH
)2
V 1/4 = 1.06× 1016( rφ
0.01
)1/4
GeV
where rφ is the tensor to scalar ratio in CMB scales [7].
• Consistency relation - r gives a test for consistency for single-field slow roll inflation
as
r = −8ntsin2∆
where sin2∆ parameterizes the ratio between the adiabatic power spectrum at
horizon-exit during inflation and the observed power spectrum giving information
about presence of multiple fields. For non-slow roll inflation, with a non-trivial cs,
it is given as
r = −ntcswhere it is seen that the non-slow roll evolution of the inflaton field is driven by a
non-canonical kinetic term.
• Inflaton Field Excursion - The inflaton field ∆φ is related to r giving the effective
number of e-folds as
Neff =
∫ N?
0
dN(r(N)
r?
) 12
where N? is the number of e-folds between the end of inflation and the horizon exit
of the CMB pivot scale. By determining, if the inflaton-field excursion was super-
planckian or not we can probe aspects of scalar field space and the UV completion
of gravity.
Constraints on Tensor to Scalar Ratio using WKB Approximation 4
• Inflationary models - The shape of the inflaton potential is controlled by the slow-
roll parameters and hence measurements of r, ns, nt, αs, αt help in constraining
different inflationary models and also removes the degeneracies between the slow-
roll parameters.
3. Alternatives for Standard Slow-roll approximation
There are many models proposed which give the similar observations, which doesn’t
explicitly consider slow roll approximation for inflation but gives very similar results
without the approximation [14]. For example,
3.1. Hamilton-Jacobi formulation of inflation
Hamilton-Jacobi formulation [15, 16] is a method to rewrite the equations of motion.
It considers the scalar inflaton field itself to be time-varying but it breaks down during
the oscillatory epoch that ends inflation [17, 18].
We know the equations of motion for a spatially flat universe as
H2 =1
3M2pl
(V (φ) +1
2φ2) (1)
and
φ+ 3Hφ = −dVdφ
, (2)
where V (φ) is the potential of the scalar field. By differentiating the first equation of
motion with respect to time and by substituting in the second equation of motion, we
get
2H = − φ2
M2pl
(3)
Dividing both sides by φ gives,
φ = −2M2plH
′(φ) (4)
which gives the Freidmann-equations in the first order form,
H ′(φ)2 − 3
2M2pl
H2(φ) = − 1
2M4pl
V (φ). (5)
When H(φ) is specified it gives the corresponding potential and all the other inflationary
solutions can be derived from it. For example,
H(φ) ∝ exp(−√
1
2p
φ
Mpl
)(6)
gives the Power-law inflation. The slow-roll parameters can be written as,
εH = 3φ2
2/(V +
φ2
2
)= −dlnH
dlna(7)
Constraints on Tensor to Scalar Ratio using WKB Approximation 5
ηH = − φ
Hφ= −dlnφ
dlna(8)
and inflation is precisely given as,
a > 0⇐⇒ εH < 1. (9)
3.2. Perturbed Power law inflation
With the standard Power law model the scale factor is given as, a(t) = tp where p > 1
and the slow-roll parameters are given as ε = −η = 1p. The Power-law also states that
[19] - [22],
(a) Expansion is uniformly accelerated ie ε = constant
(b) The perturbations in the inflaton field is similar to that for a massless scalar field
m2eff = 0.
The power spectrum is given as Pk = Akns−1 and spectral indexes are ns and nt = ns−1
which are constant values.
The perturbed Power law [23, 24] can be given as a slight deviation from the scale
invariant spectra. Here, the inflaton potential is estimated from the ‘reduced Hamilton
-Jacobi equation’. The corresponding slow-roll parameters are given as, ε =M2
pl
4π
(H′
H
)2
and η = −M2pl
4π
(H′′
H
). The spectral indices are then,
nt = ns − 1 =−2ε
1− ε(10)
where ε = (∂lnH∂φ
)2.
The Perturbed power-law is taken for small deviations from the uniform acceleration
in terms of ∂2lnH∂φ2
. The scalar perturbation spectrum is then obtained as,
Ps(k) =A(µT , µS)
2π
(H2π
)2 1
ε(η)(11)
and the corresponding tensor power-spectrum is,
Pt(k) =8A(µT , µS)
2π
(H2π
)2
(12)
where A(x, y) = 4yΓ2(y)(x−1/2)2y−1 , Γ(y) is the Euler gamma function where µT → µ and
µs =(µ2 − (µ− 1/2)2m
2eff
H2
)1/4
.
4. WKB approximation
The WKB approximation is a method used to get approximate solutions to linear
differential equations. Here it is applied on the Radial wave equations. It was proved
that the Balmer formula for Hydrogen atom could be derived from correct application
of the WKB approximation. Hence the method was used for cosmological perturbations
Constraints on Tensor to Scalar Ratio using WKB Approximation 6
which had an analogy with the Hydrogen atom formalism. The use of this method also
gives the advantage that it can be used for both subhorizon and the superhorizon scales
and it is also an alternate method than the standard slow-roll approximation.
The non-singular solutions with initial dense de-sitter phase, first mentioned by
Starobinsky gave the tensor perturbation formulation [11]. Mukhanov and Chibisov
proposed that the spectrum of quantum fluctuation in the non-singular solution could
evolve into the present universe and derived the scalar perturbation equation [25],
followed by works of Hawking [26] and Starobinsky [27]. These scalar and tensor
perturbations formulations are reduced to a single variable [28] denoted by µS or µT ,
and the equation is similar to the Schrodinger equation [29].
d2
dη2µ+ (k2 − U(η))µ =
d2µ
dη2+ ω2(η)µ = 0 (13)
where U(η) is the potential. Then the corresponding power spectra and the spectral
indices are given as,
Ps =k3
8π2
∣∣∣µszs
∣∣∣2, PT =2k3
π2
∣∣∣µTzT
∣∣∣2 (14)
where zs = a√−aa′′a′
and zT = a. The spectral indices being,
ns − 1 =∣∣∣dlnPsdlnk
∣∣∣k=k∗
, nT =∣∣∣dlnPTdlnk
∣∣∣k=k∗
. (15)
The WKB spectra is taken as the solution for the equation,
µ′′WKB(k, η) + [ω2(k, η)−Q(k, η)]µWKB = 0 (16)
where
Q(k, η) =3
4
(ω′)2
ω− ω′′
2ω. (17)
The mode function µWKB is an approximation to the actual mode function in the limit,∣∣∣ Qω2
∣∣∣ < 1 (18)
For application of the WKB approximation, the variables are transformed to
x = ln(Hak
), u = (1− ε1)
12 e
x2µ (19)
which in turn gives the Power spectra to be,
Ps =H2
πε1M2pl
( k
aH
)3 e2ψs
(1− ε1)|ωs|(20)
PT =16H2
πM2pl
( k
aH
)3 e2ψT
(1− ε1)|ωT |(21)
Here we have seen a brief on the general cosmological perturbation equations and
the application of WKB approximation on them as was given in the work by J. Martin
and D. Schwarz (2003) [30] and the derivations of the power spectra for the scalar and
the tensor perturbations. Further we use the above method for Power law inflation as
the model is exactly solvable and hence comparable with the observation data.
Constraints on Tensor to Scalar Ratio using WKB Approximation 7
5. Application of WKB approximation on Power law inflationary model
For Power law inflation, the scale factor is
a(η) = lo|η|1+β (22)
where β ≤ −2 and lo is the Hubble’s radius which is a constant when β = −2. The
corresponding Horizon flow functions are given as,
εn+1 =dln|εn|dN
(23)
for n ≥ 0. Inflation takes place when ε1 < 1. The exact power spectrum is given as
Pζ =l2pll20
f(β)k2β+4
πε1(24)
Ph =l2pll20
16f(β)k2β+4
π(25)
where lpl = m−1pl is the Planck Length and the function f(β) is given by
f(β) =1
π
[Γ(−β − 12)
2β+1
]2
(26)
where Γ is the Euler integral of the second kind.
When f(β = −2) = 1, this becomes as a special case because when β = −2 the
expression of the scalar spectrum blows up.
By applying the WKB approximation the effective frequency is given as,
ω2(η) = k2 − β(β + 1)/η2 (27)
and | Qω2 | = 1
4β(β+1). As for inflation Q
ω2 1, when β = −2 the ratio goes to 1/8 hence
satisfying the condition.
By applying the WKB approximation we get the power spectra to be
Pζ =l2pll20
1
πε1g(β)k2β+4 (28)
Ph =l2pll20
16
πg(β)k2β+4 (29)
and the function is given as,
g(β) =2e2β+1
(2β + 1)2β+2. (30)
The corresponding spectral indices are then computed to be
ns − 1 = nT = 2β + 4. (31)
The Constant-roll inflation or the Ultra-slow-roll inflationary model is also a viable
alternative to the standard slow-roll approximation. In this model, the inflationary
scenario is investigated with constant rate of roll, Φ/HΦ = −3−α (remains a constant)
[31]. The model gives experimentally viable results for α . −3. The parameter
constrains on the constant roll model and comparison with observations is explored by
Motohashi and Starobinsky [32] which gives a permitted range of r values as r . 0.07
to r 10−3.
Constraints on Tensor to Scalar Ratio using WKB Approximation 8
6. WKB Power Spectra and Tensor to scalar ratio(r)
By using the WKB approximation, we first see that the effective frequencies in Eq. (27)
are given as,
ω2s = k2η2 −
(9
4+ 3ε1 +
3
2ε2
)(32)
ω2T = k2η2 −
(9
4+ 3ε1
)(33)
Then corresponding WKB spectra are derived from solving Eq. (28) & (29) with
the above frequencies and we get the scalar power spectrum as,
Pζ =H2
πε1m2pl
(18e−3)[1− 2(D+ 1)ε1−Dε2− (2ε1 + ε2)ln( kk∗
)+ O(ε2n)](34)
and the corresponding tensor power spectrum as,
Ph =16H2
πm2pl
(18e−3)[1− 2(D + 1)ε1 − 2ε1ln( kk∗
)+ O(ε2n)] (35)
with D ≡ 13− ln3 ≈ −0.765.
The Spectral indices calculated from above equations are in accordance with the
spectral indices given by slow-roll inflation but it gives the next higher order terms
which increases it’s accuracy. The spectral indices are obtained as,
ns−1 =−2ε11− ε1
− ε21− ε1
+−2ε1 − ε2
[1− 2(D + 1)ε1 −Dε2 − (2ε1 + ε2)ln( kk∗
)]+O(εn)(36)
nT =−2ε11− ε1
+−2ε1
[1− 2(D + 1)ε1 − (2ε1)ln( kk∗
)]+O(εn) (37)
With Slow-roll approximation the power spectra are [33]
Pζ =H2
πε1m2pl
[1− 2(C + 1)ε1 − Cε2 − (2ε1 + ε2)ln( kk∗
)+ O(ε2n)] (38)
and the tensor powerspectrum,
Ph =16H2
πm2pl
[1− 2(C + 1)ε1 − 2ε1ln( kk∗
)+ O(ε2n)] (39)
where C=γE + ln2 − 2 ' -0.7296 , γE '0.5772 which is the Euler constant. All the
quantities are calculated at η∗ which is when k∗ = aH(N∗). This scale k∗ is called the
Pivot scale. The spectral indices with the slow-roll approximation are given as,
ns − 1 = −2ε1 − ε2, nT = −2ε1 (40)
From this we could readily see that the spectral indices of the WKB approximation
reduces to the spectral indices of the slow-roll approximation.
The scalar and tensor power spectra are represented by using the Power law form.
For which the expressions for Hubble’s constant is
H = − H
H2=
H ′
aH2= (1 + β)η−(β+2)/lo (41)
Constraints on Tensor to Scalar Ratio using WKB Approximation 9
and ε2 = 0. The power-spectra are analytically represented and by setting the pivot
scale at k∗ = 0.05Mpc−1.
The plots show the scale-invariant scalar and tensor power spectra with the slow-roll
approximation and with the WKB approximation. The power-spectra are calculated for
the β = −2.005.
Figure 1. Scalar power-spectra (1)with slow-roll approximation and (2)with
WKB approximation
It is clear that the power spectra are scale invariant with both approximation
techniques and the WKB approximation spectra shows a slight variation in the
amplitude which can be compensated by changing the amplitude terms in Eq. (34)
and (35). The corresponding spectral indices are calculated for these spectra. With
the WKB approximation it is ns − 1 = nT = 3.12177276e − 04 and with the slow roll
approximation ns − 1 = nT = 3.08732253e− 04.
The application of WKB approximation helps in determining the power spectra
even when the Slow-roll approximation is invalid and hence gives us a new frontier for
viewing inflation. The added advantage is that it gives solutions with higher accuracy
and higher order terms which is a necessity to match with the very precise observations
of WMAP and Planck.
We use CosmoMC [34], to establish the constrains on the parameters and to identify
the maximum likelihood region in the parameter space. The primordial power-spectrum
in CAMB of CosmoMC is changed to the WKB approximation power spectrum. As even
a slight change can produce large deviations in the CMB sky, we have constrained the
WKB approximation parameter to β = −2.005. The modified power spectrum is used
to then generate the CMB angular power spectrum with flat priors and using the whole
Constraints on Tensor to Scalar Ratio using WKB Approximation 10
Figure 2. Tensor power spectra (1)with slow roll approximation and (2)with WKB
approximation
base-set of parameters with power law ΛCDM model.
We stay with the minimal six parameter spatially flat ΛCDM cosmological model
as our base model with base parameters as baryon density today Ωbh2, cold dark matter
density today Ωch2, amplitude of scalar power spectrum As, spectral index of scalar
power spectrum ns, tensor-to-scalar ratio r and the tilt of tensor power spectrum ntbut with WKB primordial power spectrum. This model has four free non-primordial
cosmological parameters same as the power-law baseline i.e ωb, ωc, θMC and τ . Here
we try to estimate the impact of primordial power spectrum with higher-order terms of
Hubble-flow functions, which are analogous to slow-roll parameters with emphasis on
the deviations in tensor-to-scalar ratio. We also take to consideration three neutrinos
species, with two mass-less states and a single massive neutrino of mass mν = 0.06eV.
The Planck likelihood code† (PLC/clik) and parameter chains are available from
the Planck Legacy Archive. The Planck lensing and Bicep-Keck-Planck likelihoods are
included with the cosmomc.
† 2018 Planck CMB likelihoods (except for lensing) are not published yet. The present discussion refer
to the 2015 likelihoods.
Constraints on Tensor to Scalar Ratio using WKB Approximation 11
0.022 0.023
b h 2
0.11 0.115 0.12 0.125
c h 2
1.04 1.042
100 MC
0.05 0.1 0.15 3 3.1 3.2
ln (10 10 A s )
0.94 0.96 0.98 1
n s
0.78 0.8 0.82 0.84 0.86
8
0.99 1 1.01
y cal
0.996 0.998 1
c 100
2 4 6 8 10
A tSZ143
200 300 400
A PS100
64 66 68 70 72
H 0
0.66 0.7 0.74 0.26 0.3 0.34
m
0.135 0.14 0.145 0.15
m h 2
0.094 0.096 0.098
m h 3
5 10 15
z re
0 0.1 0.2
r 0.01
0.94 0.96 0.98 1
n s,0.002
0 0.05 0.1
r L=10
2 2.2 2.4
10 9 A s
0 0.2 0.4
10 9 A t
1.85 1.9 1.95
10 9 A s e -2
0 0.2 0.4
10 9 A t e -2
0.245 0.2454 0.2458
Y P
13.7 13.8 13.9
Age / Gyr
1088 1090 1092
z *
143 144 145 146 147
r *
1.04 1.042
*
0.07 0.072 0.074
r drag /D V (0.57)
0 10 20
2 prior
40 45 50 55
2 BKPLANCK
10 15 20 25
2 lensing
750 760 770 780
2 plik
800 820 840
2 CMB
Figure 3. 1D marginalized posterior distribution of the inflationary parameters
and cosmological parameters with WKB approximation
Constraints on Tensor to Scalar Ratio using WKB Approximation 12
The limits and marginalized densities given in Fig. 3 are generated using the getdist
package which is part of CosmoMC, it uses the baseline likelihood which is a hybrid, by
connecting together a low-multipole likelihood with the Gaussian likelihood constructed
from the higher multipoles. We study the Bayesian factors of the model with respect
to the power-law base model and the χ2eff values with power-law model. As it is clear
the WKB power spectrum is scale invariant and the parameters are in the 95% CL of
Planck results.
0.05 0.1 0.15 0.2r
0 0.05 0.1 0.15 0.2r0.002
Figure 4. 1D marginalized posterior distribution of the Tensor to Scalar ratio
r with WKB approximation
By combining Planck temperature, low- polarization and lensing 2018 Planck results
give r0.002 < 0.10 (95% CL, Planck TT+lowE+lensing) [10]. This constraint slightly
improves on the corresponding Planck 2015, 95% CL bound, i.e., r0.002 < 0.11 [35] which
is consistent with the B-mode polarization constraint r < 0.12 (95 % CL), obtained
from a joint analysis of the BICEP2/Keck Array and Planck data. With the WKB
approximation, the upper limit on r is obtained as r < 0.1109 and r0.002 < 0.1055
(BKPlanck†+ lensing). This gives us a window into higher order corrections and better
fit to the Planck results. It is seen that the WKB best fit value of the spectral index is
ns = 0.9657 with Ho = 67.64 which are in the 95% CL of Planck observations.
7. Conclusion
Present work studied the scalar and tensor power spectra with the WKB approximation
and constrained the value of the WKB parameter β. The WKB approximation helps
in determing the power spectra even when the Slow-roll approximation is invalid, gives
the higher order corrections and the precise details of Power law model which will give
insights into paramount questions in the Inflationary model of Universe. The spectral
index values and spectra generated with WKB approximation are compared with the
† BICEP2/Keck Array and 2015 Planck joint analysis
Constraints on Tensor to Scalar Ratio using WKB Approximation 13
observation data and we see that the WKB approximated power spectra could become
a viable alternative for Slow-roll approximation. Running Monte-Carlo chains with
CosmoMC and generating the plots with Getdist, the tensor to scalar ratio(r) is also
analysed by constraining the cosmological parameters with WKB approximation, setting
an upper limit at r < 0.1109. Here r is defined at the pivot scale k∗ = 0.05Mpc−1. We
also report the bounds on r0.002, the tensor to scalar ratio at k∗ = 0.002Mpc−1 as
r0.002 < 0.1055 which is consistent within 1σ with the Planck results.
8. Acknowledgments
We acknowledge the use of high performance computing system at IUCAA. MJ
acknowledges the support from U.G.C. Major research project grant, F.No. − 43 −523/2014(SR) and the Associateship of IUCAA. AA acknowledges the Junior Research
Fellowship from University Grants Commission (UGC), India. We thank the referee for
the valuable suggestions and comments.
References
[1] A. A. Starobinsky, Phys. Lett. B 91, 99 (1980)
[2] A. H. Guth, Phys. Rev. D. 23 347 (1981).
[3] A. D. Linde, Phys. Lett. B 108, 389 (1982).
[4] George Smoot, AIP Conf. Proc. 476, 1 (1999).
[5] H. C. Chiang, Astrophys.J. 711, 1123 (2010).
[6] K. T. Story, Astrophys.J. 779, 86 (2013).
[7] Daniel Baumann, AIP Conf.Proc. 1141, 10 (2009).
[8] Keck Array and BICEP2 Collaborations: P. A. R Ade et al, Phys. Rev. Lett. 116, 031302 (2016);
BICEP2/Keck and Planck Collaborations: P. A. R Ade et al, Phys. Rev. Lett. 114, 101301
(2015).
[9] Planck Collaboration: N. Agranim et. al, arXiv:1807.06209 (2018).
[10] Planck Collaboration: P.A.R. Ade et al, arXiv:1807.06211 (2018).
[11] A. A. Starobinsky, JETP Lett. 30, 682 (1979).
[12] Mike S. Wang, Primordial Gravitational Waves from Cosmic Inflation, Mathematical Tripos Part
III Essay 75 (2017).
[13] A. R. Liddle, An Introduction to Cosmological Inflation, arXiv:astro-ph/9901124 (1999).
[14] J. Martin and D. J Schwarz, Phys. Rev. D 62 103520 (2000).
[15] A. G. Muslimov, Class.Quant. Grav. 7, 231 (1990).
[16] D. S. Salopek and J. R. Bond, Phys. Rev. D 42, 3936 (1990).
[17] A. R. Liddle and D. H. Lyth, Cosmological Inflation and Large-Scale Structure (Cambridge
University Press, 2000).
[18] Minu Joy, V. Sahni and A. A. Starobinsky, Phys. Rev. D 77 023514 (2008).
[19] F. Lucchin and S. Matarrese, Power-law inflation, Phys. Rev. D 32, 1316 (1985).
[20] V. Sahni, The Energy Density of Relic Gravity Waves From Inflation, Phys. Rev. D 42, 453,
(1990).
[21] T. Souradeep and V. Sahni, Density perturbations, gravity waves and the cosmic microwave
background, Mod. Phys. Lett. A7, 354, (1992).
[22] T. Souradeep, Quantum phenomena in gravitation and cosmology - CMB anisotropy & the physics
of the early universe, Ph.D Thesis, IUCAA, (1995).
[23] Minu Joy and T. Souradeep, JCAP 1102 016 (2011).
Constraints on Tensor to Scalar Ratio using WKB Approximation 14
[24] S. Mukherjee, S. Das, Minu Joy and T. Souradeep, JCAP 01 043 (2015).
[25] V. F. Mukhanov and G. V. Chibisov , JETP Lett. 33, 532 (1981).
[26] S. W. Hawking, Phys. Lett. B 115, 295 (1982).
[27] A. A. Starobinsky, Phys, Lett. B 117, 175 (1982).
[28] A. H. Guth and S.-Y. Pi, Physical Review Letters 49 1110 (1982).
[29] E. Schrodinger, Phys. Rev. 28, 1049 (1926).
[30] J. Martin and D.J Schwarz, Phys. Rev. D 67 083512 (2003).
[31] H. Motohashi, A. A. Starobinsky and J. Yokoyama, JCAP 1509, 018 (2015).
[32] H. Motohashi , A. A. Starobinsky, Europhys. Lett. 117, 39001 (2017).
[33] E.D Stewart and D.H Lyth, Phys. Lett. B 302 171 (1993).
[34] http://cosmologist.info/cosmomc/
A. Lewis and S. Bridle, Phys. Rev. D 66, 103511 (2002).
[35] Planck Collaboration: P.A.R. Ade et al, Astron. Astrophys. 594 A20 (2016).