1: \def\CTeXPreproc{Created by ctex v0.2.4c, don't edit!}
2: \documentclass[aps,twocolumn,prd,nofootinbib]{revtex4}
3: \usepackage{amsmath}
4: \usepackage{graphicx}
5: \usepackage{dcolumn}
6: \usepackage{bm}
7: \usepackage{amssymb}
8: \usepackage{latexsym}
9:
10:
11: \bibliographystyle{unsrt}
12:
13: \begin{document}
14:
15: \title{Probing Noncommutativity with Inflationary Gravitational Waves}
16:
17: \author{Yi-Fu Cai$^a$\footnote{Email Address: caiyf@mail.ihep.ac.cn} and Yun-Song Piao$^{b}$\footnote{Email Address: yspiao@gucas.ac.cn}}
18: \affiliation{${}^a$Institute of High Energy Physics, Chinese
19: Academy of Sciences, P.O. Box 918-4, Beijing 100049, P. R. China}
20: \affiliation{${}^b$College of Physical Sciences, Graduate School
21: of Chinese Academy of Sciences, Beijing 100049, China}
22:
23: \begin{abstract}
24:
25: In this paper we study the behaviour of gravitational wave
26: background (GWB) generated during inflation in the noncommutative
27: field approach. From this approach we derive one additional term,
28: and then we find that the dispersion relation of the gravitational
29: wave would be modified and the primordial gravitational wave would
30: obtain an effective mass. Therefore it breaks local Lorentz
31: symmetry. Moreover, this additional term suppresses the energy
32: spectrum of the GWB greatly at low energy scale where the wave
33: length is near the current horizon. Due to this, a sharp peak is
34: formed in the energy spectrum at the low frequency band. This peak
35: should be a key criterion in detecting the spacetime
36: noncommutativity and a critical test for Lorentz symmetry breaking
37: in local field theory.
38:
39: \end{abstract}
40:
41: \maketitle
42:
43: \hskip 1.6cm PACS number(s): {98.80.Cq, 04.30.-w} \vskip 0.4cm
44:
45:
46:
47:
48: \section{Introduction}
49:
50: Under the standard model of cosmology plus the theory of
51: inflation\cite{Guth81,Steinhardt82,Linde82}, it is very natural to
52: predict the existence of the gravitational wave background. Scalar
53: type and tensor type primordial perturbations are generated during
54: inflation. The primordial scalar perturbations can be tested in
55: current observations of cosmic microwave
56: background\cite{CMBobserve,WMAP}. And they provide seeds for the
57: large scale structure(LSS), which then gradually forms today's
58: galaxies\cite{Mukhanov81,Guth82,Hawking82,Starobinsky82,Bardeen83}.
59: The tensor perturbations escape the horizon during inflation, and
60: then they keep conserved and form the relic GWB which carries the
61: information of the very early universe. There have been a number
62: of detectors operating for the signals of primordial GWB, e.g. Big
63: Bang Observer (BBO)\cite{BBO}, Planck\cite{CMBnextobserve}.
64: Besides, there have been evidences from the indirect detections of
65: the next generation of CMB observations, see related
66: analysis\cite{VPJ,Boyle06,Smith06}. The basic mechanism for the
67: generation of primordial GWB in cosmology has been discussed in
68: Refs. \cite{Grishchuk75,Allen88}. See \cite{Lyth93,Starobinski79}
69: for the gravitational waves generated during the epoch of
70: inflation; see \cite{Gasperini93} for that in Pre Big Bang
71: scenario; see \cite{Boyle04} for that in cyclic universe; see
72: \cite{piao06} for that from phantom super-inflation (also see
73: \cite{piao03,pin-others}).
74:
75: Generally speaking, the GWB was generated in the epoch when the
76: energy scale of our universe is extremely high. Thus we should
77: take into consideration on more fundamental theories in logic,
78: namely, the string theory. In string theory it is commonly
79: realized that the spacetime coordinates of a Dp-brane is
80: noncommutative under certain external fields, and this can be
81: described by the language of noncommutative
82: geometry\cite{Douglas97,Seiberg99,Bigatti00,Alekseev00,Chu02}.
83: Accordingly, there are a lot of new properties, opening up new
84: challenges for current theories and experiments. For example,
85: causality is possibly violated as a consequence of the
86: noncommutative behaviour (see Refs. \cite{Seiberg00,Gomis00}), and
87: thus the Lorentz invariance\cite{Carroll01,Carson02,Chaichian04}
88: is also probably broken. When the noncommutativity is taken into
89: account during the very early universe, it would provide large
90: corrections to inflation, the primordial
91: fluctuations\cite{Brandenberger00,HuangLi,Cai04,HuangZhang},
92: black-body spectra of CMB\cite{Fatollahi}, and also primordial
93: magnetic field\cite{BY}. Furthermore, the paper \cite{Carmona02}
94: suggested another mechanism, named noncommutative field approach,
95: which can also lead to the Lorentz invariance breaking under
96: noncommutative fields while leaving spacetime commutative. The key
97: idea of this mechanism is to construct an effective action which
98: contains noncommutative terms partly obtained from noncommutative
99: geometry. Later, this noncommutative field approach has been
100: generalized to various models and has made many interesting
101: predictions (see Refs.
102: \cite{Gamboa05,Nascimento06,Arias06,Ferrari06}). If Lorentz
103: symmetry is indeed broken in our world, it would result in the
104: breaking of the CPT symmetry, which is possible to be detected in
105: experiments.
106: % and baryogenesis from new origin\cite{Mingzhe06},
107: %also it would violate the dominant energy condition(see related
108: %works in \cite{Violatedec}).
109: The paper \cite{Feng06} has already found some evidences to support
110: that cosmic CPT symmetry may not be conserved from the analysis of
111: data fitting for cosmological observations.
112:
113: In our paper, using the noncommutative field approach, we
114: investigate the gravitational wave background during inflation.
115: The outline of this paper is as follows. In section II, we
116: introduce the additional term by applying the noncommutative field
117: approach to the equation of motion for the gravitational
118: perturbations in the flat Friedmann-Robertson-Walker(FRW)
119: background. Then we quantize the noncommutative tensor
120: perturbations and discuss the consistent initial conditions for
121: primordial gravitational waves with the noncommutative terms.
122: After that, we derive the classical solution of primordial tensor
123: perturbation with noncommutativity, and analyze the behaviour of
124: the primordial power spectrum and the corresponding spectral
125: index. In section III, from considering the influence of possible
126: factors on the energy spectrum, we discuss the transfer function
127: for the noncommutative GWB in order to relate the current GWB we
128: may probe with the primordial one. Finally, we give the
129: implications of the results from calculating the noncommutative
130: GWB in section IV, and compare them with the future development of
131: the cosmological observations. Then we summarize the fruitful
132: behaviors of GWB possessing noncommutativity which may be detected
133: in the future probe.
134:
135: In this paper, we will see that, the dispersion relation of the
136: primordial gravitational wave will be modified, and its power
137: spectrum will exhibit difference from the normal one around the
138: CMB scale due to the noncommutative term. Furthermore, we can
139: obtain the same viewpoint from the relationship between the tensor
140: spectral index and its frequency. From the modified primordial
141: power spectrum, we can show that there will be a sharp peak in the
142: CMB scale of the energy spectrum observed today. In order to add
143: the contribution of the transfer function, we discuss some leading
144: corrections to the primordial GWB during the evolution of the
145: universe. Among these corrections, we mainly consider the
146: suppression of the GWB from the redshift, the impression of the
147: background equation of state of the universe when the GWB
148: re-enters the horizon, and the damping effects from the
149: anisotropic stress generated from some particles' freely
150: streaming\cite{Weinberg03,Pritchard04,Bashinsky,Dicus05}. After
151: considering all these contributions, we can clearly find that the
152: effects of noncommutative terms still play a significant role in
153: the low frequency range. So the results and predictions in the
154: above sections are important for future observations.
155:
156:
157:
158:
159: \section{Noncommutative tensor perturbations during inflation}
160:
161:
162: \subsection{Bases and Conventions}
163:
164: In this section, we take a brief review of the standard theory of
165: tensor perturbation during inflation(for more details, see the
166: review article Ref. \cite{Brandenberger92}). In the flat FRW
167: background, the tensor perturbation is given in the metric,
168: \begin{eqnarray}
169: ds^2=a(\tau)^2[-d\tau^2+(\delta_{ij}+\bar h_{ij})dx^idx^j]~,
170: \end{eqnarray}
171: where $^{(0)}g_{\mu\nu}=diag(-a^2,a^2,a^2,a^2)$ is the background
172: metric, $\tau$ is the conformal time, $a(\tau)$ is the scale
173: factor, and the Latin indexes represent spatial coordinates. Here
174: the perturbation $\bar h_{ij}$ satisfies the following
175: constraints:
176: \begin{eqnarray}
177: \bar h_{ij}=\bar h_{ji}~;~\bar h_{ii}=0~;~\bar h_{ij,j}=0~.
178: \end{eqnarray}
179: Therefore, $\bar h_{ij}$ only have two degrees of freedom
180: corresponding to two polarizations of gravitational waves.
181:
182: To start from Einstein's field equation, one can deduce the action
183: of free tensor perturbation as follows:
184: \begin{eqnarray}
185: ^{(2)}{\cal S}_g=\int d\tau d^3x \frac{1}{64\pi G}a^2[\bar
186: h_{ij}'^2-(\partial_l{\bar h_{ij}})^2]~,
187: \end{eqnarray}
188: where prime represents the derivative with respect to the conformal
189: time. The interaction part of the action with other matter sources
190: is of the form
191: \begin{eqnarray}
192: ^{(2)}{\cal S}_m=\int d\tau d^3x\frac{1}{2}a^4\sigma_{ij} \bar
193: h_{ij}~,
194: \end{eqnarray}
195: Here $\sigma_{ij}$ is the anisotropic part of the stress tensor,
196: constructed by the spatial components of the perturbed
197: energy-momentum tensor ${T^i}_j$,
198: \begin{eqnarray}
199: \sigma_{ij}=T^i_j-^{(0)}p\delta^i_j~.
200: \end{eqnarray}
201: Then one can derive the general equation of motion for tensor
202: perturbation:
203: \begin{eqnarray}
204: \bar h_{ij}''+2\frac{a'}{a}\bar h_{ij}'-{\bf \nabla}^{2}\bar
205: h_{ij}-16\pi G a^{2}\sigma_{ij}=0~.
206: \end{eqnarray}
207:
208: The Fourier transformation of the tensor perturbation and the
209: anisotropic stress tensor takes the form,
210: \begin{eqnarray}
211: \bar h_{ij}(\tau, {\bf x})=\sqrt{16\pi
212: G}\int\frac{d^3k}{(2\pi)^\frac{3}{2}}H_{ij}(\tau, {\bf k})e^{i{\bf
213: k}{\bf x}}~,\\
214: \sigma_{ij}(\tau, {\bf x})=\sqrt{16\pi
215: G}\int\frac{d^3k}{(2\pi)^\frac{3}{2}}\Sigma_{ij}(\tau, {\bf
216: k})e^{i{\bf k}{\bf x}}~,
217: \end{eqnarray}
218: where we leave the Latin indices in order to keep the following
219: progresses to be general. In cosmology, what we care about is the
220: distribution of the spectra of gravitational waves and the
221: corresponding spectral index. Based on the above formalism, the
222: tensor power spectrum can be written as,
223: \begin{eqnarray}\label{tensor power}
224: P_T(k,\tau)&\equiv\frac{d\langle0|\bar{h}_{ij}^{2}|0\rangle}{d\,{\rm
225: ln}\,k}=32\pi G\frac{k^3}{(2\pi)^2}|H_{ij}(\tau, {\bf k})|^2~,
226: \end{eqnarray}
227: and the definition of tensor spectral index $n_T$ is given by
228: \begin{eqnarray}
229: n_T\equiv\frac{d\,{\rm ln}\,P_T}{d\,{\rm ln}\,k}~.
230: \end{eqnarray}
231:
232: The GWB we observed today is characterized by the energy spectrum,
233: \begin{eqnarray}\label{tensor energy}
234: \Omega_{GW}(k,
235: \tau)\equiv\frac{1}{\rho_c(\tau)}\frac{d\langle0|\rho_{GW}(\tau)|0\rangle}{d\,{\rm
236: ln}\,k}~,
237: \end{eqnarray}
238: where $\rho_{GW}(\tau)$ indicates the energy density of
239: gravitational waves, and the parameter $\rho_c(\tau)$ is the
240: critical density of the universe. Make use of the Friedmann equation
241: \begin{eqnarray}
242: H^2(\tau)=\frac{8\pi G}{3}\rho_c(\tau)~,
243: \end{eqnarray}
244: the energy spectrum of GWB can be written as,
245: \begin{eqnarray}\label{approximate tensor energy0}
246: \Omega_{GW}(k,\tau)=\frac{8\pi
247: G}{3H^2(\tau)}\frac{k^3}{2(2\pi)^2}\frac{1}{a^2}\left(|H_{ij}'|^2+k^2|H_{ij}|^2\right).
248: \end{eqnarray}
249: In respect that the GWB we observed has already re-entered the
250: horizon, its mode should oscillate in the form of sinusoidal
251: function. Accordingly, we can deduce the relation between the power
252: spectrum and the energy spectrum we care about as follows
253: \begin{eqnarray}\label{approximate tensor energy}
254: \Omega_{GW}(k,\tau)\simeq\frac{1}{12}\frac{k^2}{a^2(\tau)H^2(\tau)}P_T(k,
255: \tau)~.
256: \end{eqnarray}
257: Note that, the equations (\ref{approximate tensor energy0}) and
258: (\ref{approximate tensor energy}) can be extended into the case of
259: noncommutative gravitational waves because in the framework of
260: noncommutative GWB the Hamiltonian is not modified, we will see the
261: detailed analysis in the next section.
262:
263:
264: \subsection{Quantization and Noncommutativity}
265:
266: In this section, we will quantize the tensor perturbation, and
267: then introduce the noncommutative term in the canonical
268: commutation relation. Based on this, we will obtain a modified
269: theory of gravity in weak gravity approximation and finally
270: provide the equation of motion of primordial gravitational waves
271: in this theory. Since the action of tensor perturbation in
272: noncommutative field approach will break the Lorentz symmetry in
273: local Minkowski spacetime which can be exhibited in the dispersion
274: relation of gravitons\cite{Ferrari06}, we can see the
275: corresponding behaviour of gravitons in the the flat FRW universe
276: background.
277:
278: Firstly, for simplicity, we redefine the tensor perturbation as
279: follows
280: \begin{eqnarray}\label{tensor redefine}
281: H_{ij}(\tau, {\bf k})\equiv\frac{h_{ij}(\tau, {\bf k})}{a(\tau)}~.
282: \end{eqnarray}
283: Therefore, in the momentum space the action of tensor perturbation
284: including the minimal coupling with matter is given by
285: \begin{eqnarray}
286: ^{(2)}{\cal S}&=&\int
287: d\tau d^3k~[\frac{1}{4}(h_{ij}'^2+\frac{a''}{a}h_{ij}^2-k^2h_{ij}^2)\nonumber\\
288: &+&8\pi Ga^3\Sigma_{ij}h_{ij}]~,
289: \end{eqnarray}
290: with the equation of motion,
291: \begin{eqnarray}\label{old h equation}
292: h_{ij}''-\frac{a''}{a}h_{ij}+k^2h_{ij}-16\pi Ga^3\Sigma_{ij}=0~.
293: \end{eqnarray}
294: Secondly, from the standard canonical quantization, the conjugate
295: momentum can be written as,
296: \begin{eqnarray}
297: p_{ij}=\frac{\delta^{(2)}{\cal S}}{\delta
298: h_{ij}'}=\frac{1}{2}h_{ij}'~,
299: \end{eqnarray}
300: and the canonical Hamiltonian density in Fourier space is given by
301: \begin{eqnarray}
302: {\cal H}&=&p_{ij}h_{ij}'-{\cal L}\nonumber\\
303: &=&\frac{1}{4}p_{ij}^2-\frac{1}{4}\frac{a''}{a}h_{ij}^2+\frac{1}{4}k^2h_{ij}^2-8\pi
304: Ga^3\Sigma_{ij}h_{ij}~.
305: \end{eqnarray}
306: Converting the fields $h_{ij}$ and the conjugated momenta $p_{ij}$
307: into operators, we obtain the equal-time commutation relations from
308: the Poisson algebra,
309: \begin{eqnarray}\label{commutation relation1}
310: [h_{ij}(\tau, {\bf k}),h_{kl}(\tau, {\bf
311: k}')]&=&0~,\\\label{commutation relation2} [p_{ij}(\tau, {\bf
312: k}),p_{kl}(\tau, {\bf k}')]&=&0~,\\\label{commutation relation3}
313: [h_{ij}(\tau, {\bf k}),p_{kl}(\tau, {\bf
314: k}')]&=&\frac{1}{2}(\delta_{ik}\delta_{jl}+\delta_{il}\delta_{jk})\nonumber\\
315: &\times&\delta^{(3)}({\bf k}-{\bf k'})~.
316: \end{eqnarray}
317: Finally, following the similar formalism in \cite{Ferrari06}, we
318: make a deformation of the commutation relation of the conjugated
319: momenta into the noncommutative form
320: \begin{eqnarray}
321: [p_{ij}(\tau, {\bf k}),p_{kl}(\tau, {\bf
322: k}')]=\alpha_{ijkl}\delta^{(3)}({\bf k}-{\bf k'})~,
323: \end{eqnarray}
324: where $\alpha_{ijkl}$ is a constant matrix. In order to preserve
325: the properties of tensor perturbation, we require that
326: $\alpha_{ijkl}$ satisfies the following relation,
327: \begin{eqnarray}
328: \alpha_{ijkl}=\alpha_{jikl}=-\alpha_{klij}.
329: \end{eqnarray}
330: This is satisfied if
331: \begin{eqnarray}
332: \alpha_{ijkl}=\alpha_{ik}\delta_{jl}+\alpha_{il}\delta_{jk}+\alpha_{jk}\delta_{il}+\alpha_{jl}\delta_{ik},
333: \end{eqnarray}
334: with $\alpha_{jk}$ a constant antisymmetric matrix. One can check
335: that $\alpha_{ijkl}$ keeps the traceless property of tensor
336: perturbation. %Meanwhile, as it is constant, the transversality is
337: %also saved.
338: Moreover, in order to simplify the calculation,
339: $\alpha_{jk}$ can be further expressed as,
340: \begin{eqnarray}
341: \alpha_{jk}=-\epsilon_{0jkl}\alpha^l,
342: \end{eqnarray}
343: where $\epsilon_{ijkl}$ is a totally antisymmetric tensor, and we
344: use the convention that $\epsilon_{0123}=-1$ and
345: $\epsilon^{0123}=+1$ in our paper.
346:
347:
348:
349: In order to recover a canonical system quantized in normal
350: commutative relations like eq. (\ref{commutation relation2}), we
351: construct a new conjugated momentum
352: $\pi_{ij}=p_{ij}-\frac{1}{2}\alpha_{ijkl}h_{kl}$ for the tensor
353: fields $h_{ij}$. Since in the derivation above we have discarded
354: the components of metric perturbations containing the time index,
355: the origin form of lagrangian is not covariant
356: % and would miss some
357: %contents when we consider non-commutativity in the process of
358: %quantization
359: . %To keep the form to be covariant,
360: %include all contributions of non-commutativity,
361: We modify the lagrangian by adding an additional term of covariant
362: form and then leave physical degrees of freedom in the following,
363: \begin{eqnarray}
364: \Delta{\cal
365: L}&=&-2\epsilon^{\rho\mu\nu\sigma}\alpha_{\sigma}h_{\nu\kappa}\partial_{\rho}h^{\kappa}_{\mu}~,\\
366: &\rightarrow&-2\alpha_0\epsilon^{0ijk}h_{kl}\partial_ih_{lj}-2\alpha_m\epsilon^{0jkm}h_{kl}h_{lj}'~,
367: \end{eqnarray}
368: where we call $\alpha_\mu$ as noncommutative parameters(NP).
369:
370: Note that after making the Fourier transformation, the term
371: containing $\alpha_0$ disappears because of the symmetry of
372: exchanging indices ``$k$" and ``$j$". Here one can check that
373: Hamiltonian still keep invariant since the term containing
374: $\alpha_m$ can be cancelled by modified conjugated momentum and
375: the term including $\alpha_0$ also disappears due to the symmetry
376: of exchanging indices. Therefore, the modified Einstein-Hilbert
377: action of tensor perturbations is given by
378: \begin{widetext}
379: \begin{eqnarray}
380: ^{(2)}{\cal S}^{new}=\int d\tau
381: d^3k~[\frac{1}{4}(h_{ij}'^2+\frac{a''}{a}h_{ij}^2-k^2h_{ij}^2)
382: -2\alpha_m\epsilon^{0jkm}h_{kl}h_{lj}' +8\pi
383: Ga^3\Sigma_{ij}h_{ij}]~,
384: \end{eqnarray}
385: \end{widetext}
386: and the corresponding equation of motion is of form
387: \begin{eqnarray}\label{h equation}
388: h_{ij}''+k^2h_{ij}-\frac{a''}{a}h_{ij}+8\alpha_m\epsilon^{0lim}h_{jl}'=16\pi
389: Ga^3\Sigma_{ij}~.
390: \end{eqnarray}
391: Compared with eq. (\ref{old h equation}), one can already find the
392: difference appeared in each one's dispersion relation roughly.
393:
394: %222
395:
396: %In order to recover a canonical system quantized in normal
397: %commutative relations like Eq. (\ref{commutation relation2}), we
398: %reconstruct the conjugate momentum as
399: %$\pi_{ij}=p_{ij}-\frac{1}{2}\alpha_{ijkl}h_{kl}$ for the tensor
400: %fields $h_{ij}$. Correspondingly, we modify the lagrangian by
401: %adding an additional term,
402: %\begin{eqnarray}
403: %\Delta{\cal L}=-2\alpha_m\epsilon^{0jkm}h_{kl}h_{lj}'~,
404: %\end{eqnarray}
405: %where we call $\alpha_m$ the noncommutative parameter(NP).
406: %Since in the derivation above we have discarded
407: %the components of metric perturbations containing the time index,
408: %the origin form of lagrangian is not covariant and would miss some
409: %contents when we consider non-commutativity in the process of
410: %quantization. To include all contributions of non-commutativity,
411: %we thus modify the lagrangian by adding an additional term of
412: %covariant form and then leave physical degrees of freedom in the
413: %following,
414: %\begin{eqnarray}
415: %\Delta{\cal
416: %L}&=&-2\epsilon^{\rho\mu\nu\sigma}\alpha_{\sigma}h_{\nu\kappa}\partial_{\rho}h^{\kappa}_{\mu}~,\\
417: %&\rightarrow&-2\alpha_0\epsilon^{0ijk}h_{kl}\partial_ih_{lj}-2\alpha_m\epsilon^{0jkm}h_{kl}h_{lj}'~,
418: %\end{eqnarray}
419: %where we call $\alpha_\mu$ as noncommutative parameters(NP).
420:
421: %Here, one can check that Hamiltonian keeps invariant after this
422: %modification. It's because the term containing $\alpha_m$ will be
423: %cancelled by modified conjugated
424: %momentum% and the term including $\alpha_0$ will disappear due to
425: %the symmetry of exchanging indices
426: %. Therefore, the modified Einstein-Hilbert action for tensor
427: %perturbation is given by
428: %\begin{widetext}
429: %\begin{eqnarray}
430: %^{(2)}{\cal S}^{new}=\int d\tau
431: %d^3k~[\frac{1}{4}(h_{ij}'^2+\frac{a''}{a}h_{ij}^2-k^2h_{ij}^2)
432: %-2\alpha_m\epsilon^{0jkm}h_{kl}h_{lj}' +8\pi
433: %Ga^3\Sigma_{ij}h_{ij}]~,
434: %\end{eqnarray}
435: %\end{widetext}
436: %and the corresponding equation of motion is,
437: %\begin{eqnarray}\label{h equation}
438: %h_{ij}''+k^2h_{ij}-\frac{a''}{a}h_{ij}+8\alpha_m\epsilon^{0lim}h_{jl}'=16\pi
439: %Ga^3\Sigma_{ij}~.
440: %\end{eqnarray}
441: %Compared with Eq. (\ref{old h equation}), one can already find the
442: %difference appeared in each one's dispersion relation roughly.
443:
444:
445:
446: \subsection{Noncommutative Tensor Perturbations During Inflation}
447:
448: In order to analyze the noncommutative tensor perturbations
449: explicitly, we need to obtain the classical solution of each mode of
450: tensor perturbations under the slow roll inflation background. Here
451: we choose NP of form: $\alpha_m=\{0,0,\alpha_3\}$ and substitute
452: this into Eq. (\ref{h equation}). Besides, we neglect the
453: anisotropic tensor fluctuations of matters and thus
454: $\Sigma_{ij}\simeq0$ here. According to the qualities of gauge
455: invariance, we can obtain only two independent modes which can be
456: expressed as
457: \begin{eqnarray}
458: v_1=\frac{1}{\sqrt{2}}(h_{11}+ih_{12})~,~~v_2=\frac{1}{\sqrt{2}}(h_{11}-ih_{12})~,
459: \end{eqnarray}
460: and in our note we call them left-handed and right-handed
461: respectively. These two modes satisfy their own equations:
462: \begin{eqnarray}\label{v1}
463: v_1''+8i\alpha_3v_1'+k^2 v_1-\frac{a''}{a}v_1=0~,\\\label{v2}
464: v_2''-8i\alpha_3v_2'+k^2 v_2-\frac{a''}{a}v_2=0~.
465: \end{eqnarray}
466:
467: To solve these two equations explicitly, we need to know their
468: initial conditions. In the noncommutative field approach the
469: dispersion relation of gravitational waves has been modified, thus
470: it seems that the initial conditions should be different from
471: those in commutative case. Thus we need to reconsider the choice
472: of initial conditions. Here we fix the initial conditions by
473: requiring the initial state to be the lowest energy state(more
474: general initial conditions of modified dispersion relation, see
475: Ref. \cite{BrandenbergerM00}).
476: %In other words, we
477: %choose the initial conditions on the vacuum of primordial
478: %fluctuations
479: Since the noncommutative term does not contribute to energy
480: density, we can use the original form appeared in Eq.
481: (\ref{approximate tensor energy0}), and then give the expression
482: of the energy density:
483: \begin{eqnarray}
484: \rho_{GW}&=&\int_0^\infty
485: dk\frac{k^2}{8\pi^2a^4}\sum_{i=1,2}[v_i'{v_{i}^*}'-\frac{a'}{a}v_i'{v_{i}^*}-\frac{a'}{a}v_i{v_{i}^*}'\nonumber\\
486: &+&\frac{a'^2}{a^2}v_i{v_{i}^*}+k^2v_i{v_{i}^*}]~,
487: \end{eqnarray}
488: where we sum 'i' over the left-handed and right-handed modes.
489: Note that, the energy density can be divided into two part, for
490: one only containing $v_1$ and for the other only containing $v_2$.
491: Accordingly, we can analyze each mode separately and then obtain
492: the initial conditions for each one. Now we define a ratio
493: $\frac{v_1'}{v_1}\equiv x(\tau)+iy(\tau)$, and for convenience we
494: convent a Wronskian ${\cal W}\equiv v_1'{v_{1}^*}-v_1{v_{1}^*}'$.
495: Then at the initial conformal time $\tau_0$ the energy density of
496: the left-handed mode is given by
497: \begin{eqnarray}
498: \rho_L=\int_0^\infty\frac{dk k^2{\cal
499: W}}{16i\pi^2a^4y_0}(x_0^2+y_0^2-2\frac{a'}{a}x_0+\frac{a'^2}{a^2}+k^2)~,
500: \end{eqnarray}
501: where $x_0=x(\tau_0)$ and $y_0=y(\tau_0)$ which are the ratio
502: parameters at the initial conformal time.
503:
504: Commonly, it is sufficient to obtain the initial conditions by
505: calculating the variation of the energy density with respect to
506: $x_0$ and $y_0$. From the vanishing variations we obtain the
507: ``initial" conditions $ x_0=\frac{a'}{a}$ and $y_0^2=k^2$.
508: However, if these two conditions are satisfied in Eq. (\ref{v1})
509: and (\ref{v2}), we have to require $\alpha_3$ to be zero which
510: violates the noncommutativity. Therefore, the lowest energy
511: density of noncommutative primordial tensor perturbations is not
512: an extremal point where the variations with respect to $x_0$ and
513: $y_0$ are zero. Instead, the lowest energy density can only be
514: chosen on the boundary value of the range we care about. Moreover,
515: this boundary value should be positive and independent of any
516: time-dependent phase. Therefore, we deduce that only if $|y_0|$
517: approaches to $k$ as closely as possible, it is satisfied that the
518: energy state is the lowest in the range we allow. Consequently, we
519: obtain the correct initial condition $y_0=-l-4\alpha_3$ for the
520: left-handed in our note; similarly, we also obtain the
521: corresponding initial condition for the right-handed.
522:
523: Thus when $|l\tau|\gg aH$, $v_1$ and $v_2$ are given by
524: \begin{eqnarray}\label{plane wave}
525: v_1\rightarrow\frac{1}{\sqrt{2l}}e^{-i(l+4\alpha_3)\tau}~,~~v_2\rightarrow\frac{1}{\sqrt{2l}}e^{-i(l-4\alpha_3)\tau}~,
526: \label{initial}\end{eqnarray} where $l$ denote the effective
527: co-moving wave numbers of the modified tensor perturbations, and
528: its form is given by
529: \begin{eqnarray}
530: \label{effective wave
531: number}l&=&(k^2+16\alpha_3^2)^{\frac{1}{2}}~.
532: \end{eqnarray}
533: Thus we see that both $v_1$ and $v_2$ display a well oscillating
534: behaviour inside of horizon, and are identical with the standard
535: primordial tensor perturbations if we neglect the noncommutative
536: terms. During inflation, in the slow roll approximation, we have the
537: relation
538: \begin{equation}
539: \frac{a''}{a}=\frac{\nu^2-\frac{1}{4}}{\tau^2}~,
540: \end{equation} where
541: $\nu=\frac{\varepsilon-3}{2(\varepsilon-1)}$ and
542: $\varepsilon\equiv-\frac{\dot H}{H^2}$ is the slow roll parameter.
543: Thus with the initial conditions (\ref{initial}), the solutions of
544: Eqs. (\ref{v1}) and (\ref{v2}) can be given as
545: \begin{eqnarray}\label{exact v1}
546: &v_1(k,
547: \tau)=\frac{\sqrt{\pi}}{2}e^{i(\nu+\frac{1}{2})\frac{\pi}{2}}(-\tau)^{\frac{1}{2}}e^{-4i\alpha_3\tau}H_{\nu}^{(1)}(-l\tau)~,\\
548: &v_2(k,
549: \tau)=\frac{\sqrt{\pi}}{2}e^{i(\nu+\frac{1}{2})\frac{\pi}{2}}(-\tau)^{\frac{1}{2}}e^{4i\alpha_3\tau}H_{\nu}^{(1)}(-l\tau)~,
550: \end{eqnarray}
551: where $H_{\nu}^{(1)}$ is the $\nu$th Hankel function of the first
552: kind. It is interesting to notice that the tensor perturbations
553: obtain an effective mass, and thus in the low energy limit the
554: effective mass would bring some brand-new phenomena. Consequently,
555: we conclude that the dispersion relation of primordial gravitational
556: waves has been modified and locally the Lorentz symmetry is broken.
557: When the modes $v_1$ and $v_2$ are out of horizon ($|l|\ll aH$), we
558: can obtain their asymptotic forms:
559: \begin{eqnarray}\label{h outof horizon1}
560: &v_1\rightarrow
561: e^{i(\nu-\frac{1}{2})\frac{\pi}{2}}2^{\nu-\frac{3}{2}}e^{-4i\alpha_3\tau}\frac{\Gamma(\nu)}{\Gamma(\frac{3}{2})}\frac{(-l\tau)^{\frac{1}{2}-\nu}}{\sqrt{2l}}~,\\
562: \label{h outof horizon2} &v_2\rightarrow
563: e^{i(\nu-\frac{1}{2})\frac{\pi}{2}}2^{\nu-\frac{3}{2}}e^{4i\alpha_3\tau}\frac{\Gamma(\nu)}{\Gamma(\frac{3}{2})}\frac{(-l\tau)^{\frac{1}{2}-\nu}}{\sqrt{2l}}~,
564: \end{eqnarray}
565: which are the solutions of primordial gravitational waves after
566: adding noncommutative terms. These solutions are used in the next
567: section.
568:
569:
570:
571: \section{What can we see from GWB today?}
572:
573: \subsection{Features of Primordial GWB}
574:
575: In this section we focus on the connections of theory and
576: observations, and hence we need to know the detailed features of
577: primordial gravitational waves. Since the noncommutative terms would
578: modify the dispersion relation of gravitational waves, it is
579: interesting to see what would be brought about in the propagations
580: and spectra of primordial tensor perturbations.
581:
582: %First of all, by using the conditions we deduced in the last section
583: %and combining eqs. (\ref{solution of v1}) and (\ref{solution of
584: %v2}), one can derive the exact form of two modes. Now we provide
585: %the results of these two modes directly:
586:
587: %Moreover, since we need to know more features of the noncommutative
588: %tensor perturbations in the epoch of inflation, it is valuable to
589: %investigate the primordial tensor power spectrum $P_T$ and the
590: %corresponding spectral index $n_T$.
591:
592: %Because in noncommutative field approach the right-handed mode of
593: %gravitational waves is independent of the left-handed one and the
594: %evolving processes of these two polarizations are different from
595: %each other. It will be more convenient to split the power spectrum
596: %into two part:
597: %\begin{eqnarray}\label{PT}
598: %P_T(k,\tau)=P_T^L(k,\tau)+P_T^R(k,\tau)~.
599: %\end{eqnarray}
600:
601: For convenience we define
602: \begin{eqnarray}
603: u_1(k,\tau)\equiv\frac{v_1(k,\tau)}{a}~,~~u_2(k,\tau)\equiv\frac{v_2(k,\tau)}{a}~.
604: \end{eqnarray}
605: Combining the definition of the power spectrum, we have the
606: expression:
607: \begin{eqnarray}\label{PT}
608: P_T(k,\tau)=32\pi
609: G\frac{k^3}{2\pi^2}(|u_1(k,\tau)|^2+|u_2(k,\tau)|^2)~.
610: %P_T^R(k,\tau)=32\pi G\frac{k^3}{2\pi^2}|u_2(k,\tau)|^2~.
611: \end{eqnarray}
612: Moreover, the power spectra of tensor perturbations at different
613: time can be connected by a parameter called transfer function.
614: Therefore, we define the transfer function as follows:
615: \begin{eqnarray}\label{PTtoday}
616: P_T(k,\tau)=T(k,\tau)P_T(k,\tau_i)~,
617: %\\\label{PTtoday2}P_T^R(k,\tau)=T^R(k,\tau)P_T^R(k,\tau_i)~,
618: \end{eqnarray}
619: where the index ($_i$) indicates the end of inflation. To
620: substitute eqs. (\ref{h outof horizon1}) and (\ref{h outof
621: horizon2}) into the definition of tensor power spectrum in Eq.
622: (\ref{PT}), we obtain the analytic solution of primordial power
623: spectrum $P_T(k)$:
624: \begin{eqnarray}\label{tensor power slow roll}
625: P_T(k)\equiv P_T(k,\tau_i)&=&64\pi G\frac{2^{2\nu-3}}{4\pi^2a_i^2}\frac{\Gamma^2(\nu)}{\Gamma^2(\frac{3}{2})}[(1-\varepsilon)a_iH_i]^{2\nu-1}\nonumber\\
626: &\times&k^3l^{-2\nu}(k)~.
627: \end{eqnarray}
628:
629: Furthermore, there have been a number of papers to investigate the
630: mass of today's graviton, and they provide an upper limit around
631: $6.395\times10^{-32}$eV
632: theoretically\cite{Gruzinov01,Cutler02,Cooray03,Jones04}. Taking
633: into account this limit, we easily find that $\alpha_3$ has to be
634: less than the frequency of $10^{-16}$Hz, whose Compton wavelength is
635: in equal order of the size of supercluster. Besides, to require that
636: the slow-roll parameter $\varepsilon$ is close to zero, we then
637: derive the form of primordial tensor power spectrum,
638: \begin{eqnarray}\label{pPT}
639: P_T(k)\simeq\frac{128\pi}{3M_{pl}^4}V_{inf}\frac{k^3}{(k^2+16\alpha_3^2)^\frac{3}{2}}~.
640: \end{eqnarray}
641: If the momentum $k$ is much larger than $\alpha_3$, one can see that
642: this primordial tensor power spectrum will return to the standard
643: form $\frac{128\pi}{3M_{pl}^4}V_{inf}$ from the equation above.
644: However, if $k$ is tuned around the value of NP, the amplitude of
645: the power spectrum starts to decrease.
646: %the first
647: %term appeared in the square brackets $[...]$ of Eq. (\ref{pPT})
648: %keeps decreasing, meanwhile, the latter one can firstly reach a
649: %maximum and then decrease as well.
650: In this theory $P_T$ would vanish when we tune $k$ to be zero
651: which is greatly different from the typical knowledge of GWB.
652:
653: \begin{figure}
654: \begin{center}
655: \includegraphics[width=3.7in]{fig1.eps}
656: \end{center}
657: \caption{The black solid curve represents the primordial power
658: spectrum of noncommutative tensor perturbations $P_T(k,\tau_i)$ at
659: the end time of inflation. The red dash curve represents the
660: primordial power spectrum without noncommutative contributions
661: while in the same inflation model. In this figure, we adopt the
662: value of NP as $\alpha_3=10^{-18}$, and the potential of inflation
663: as $V_{inf}\sim M^4$ in which $M=5\times10^{15}$Gev.}
664: \label{fig:primordial}
665: \end{figure}
666:
667: \begin{figure}
668: \begin{center}
669: \includegraphics[width=3.4in]{fig2.eps}
670: \end{center}
671: \caption{The black solid curve depicts the features of primordial
672: tensor spectral index of noncommutative tensor perturbations
673: $n_T$. As we know, the tensor spectral index in standard
674: perturbation theory would reside around zero. However in this
675: figure it does not stay at zero but runs to 3 which is strongly
676: different from the standard theory. Here we still adopt
677: $\alpha_3=10^{-18}$, and the potential of inflation as
678: $V_{inf}\sim M^4$ in which $M=5\times10^{15}$Gev.} \label{fig:nT}
679: \end{figure}
680:
681: Eventually, to make the analysis more specifical, we give a scenario
682: of primordial tensor power spectrum in the inflation of nearly
683: constant potential in Fig. \ref{fig:primordial}. Comparing the one
684: from the standard theory, our primordial tensor power spectrum shows
685: a greatly dynamical behaviour which is independent of specific
686: inflation models as shown in Fig. \ref{fig:primordial}. Note that in
687: Fig. \ref{fig:primordial}, When the value of $k$ is much larger than
688: NP, the solid line tends forwards the red dash line; when $k$
689: approaches the NP from the right side, the solid line starts
690: dropping down and finally vanishes when $k$ approaches zero. The
691: other way to see this interesting dynamical behaviour of
692: noncommutative tensor fluctuations is to analyze its spectral index
693: which is shown in Fig. \ref{fig:nT}. We find that in the range of
694: frequencies near NP, the spectral index would not reside on the
695: value $0$ but climb up to a fixed positive value which depends on
696: the model of inflation.
697:
698: \subsection{Transfer Function of Gravitational waves}
699:
700: In the above we discussed the behaviour of noncommutative tensor
701: perturbations exhibited in primordial tensor power spectrum. Now
702: what we care about is how to recognize these tensor perturbations in
703: the GWB nowadays. Since the primordial gravitational waves are
704: distributed in every frequency, once the effective co-moving wave
705: number is less than $aH$, the corresponding mode of gravitational
706: waves would escape the horizon and be frozen until it re-enters the
707: horizon. The relation between the time when tensor perturbations
708: leave the horizon and the time when they return is
709: $a_{out}H_{out}=a_{in}H_{in}$. Therefore, we have the conclusion
710: that, the earlier the perturbations escape the horizon, the later
711: they re-enter it. Moreover, once the effective co-moving wave number
712: is larger than $aH$, the perturbations begin to oscillate like plane
713: wave. In the following, we will establish the relation to relate the
714: power spectrum observed today to the primordial one. The method we
715: used here is similar to \cite{Steinhardt05}.
716:
717: In order to make clear every possible ingredient affecting the
718: evolvement of the GWB, it is suitable and reasonable to decompose
719: the transfer function into three parts as,
720: \begin{eqnarray}
721: T(k,\tau)&=&F_1F_2F_3\nonumber\\
722: &=&|\frac{\bar
723: u_{1(2)}(k,\tau)}{u_{1(2)}(k,\tau_i)}|^2|\frac{\tilde
724: u_{1(2)}(k,\tau)}{\bar
725: u_{1(2)}(k,\tau)}|^2|\frac{u_{1(2)}(k,\tau)}{\tilde
726: u_{1(2)}(k,\tau)}|^2.\nonumber\\
727: \end{eqnarray}
728: Here $u_{1(2)}(k,\tau)$ is the exact solution of Eq. (\ref{h
729: equation}); $\tilde u_{1(2)}(k,\tau)$ is an approximate solution of
730: Eq. (\ref{h equation}) by neglecting the anisotropic stress tensor
731: $\Sigma_{ij}$; and $\bar u_{1(2)}(k,\tau)$ is an even more crude
732: solution which is equal to $u_{1(2)}(k,\tau_i)$ if $k<aH$ while
733: equal to plane wave if $k>aH$. Note that those two polarizations are
734: only different on their phases which does not affect the expression
735: of the transfer function. Consequently, we only need to investigate
736: one polarization in the following.
737:
738: Firstly, from Eq. (\ref{plane wave}) one can see that after horizon
739: re-entering, gravitational waves begin to oscillate with a decaying
740: amplitude proportional to $a^{-1}(\tau)$. Therefore, from the
741: definition of $\bar u_1$ we get
742: \begin{eqnarray}
743: \bar u_1(k,\tau)= \left\{ \begin{array}{c}
744: \frac{u_{1max}(k)}{a(\tau)}\cos[l(\tau-\tau_l)+\phi_l],~~l>aH \\
745: \\
746: u_1(k,\tau_i),~~l<aH
747: \end{array} \right. \label{baru1}
748: \end{eqnarray}
749: where $\phi_l$ depends on the initial condition, $u_{1max}(k)$ is
750: the maximum of the amplitude of oscillation, and $\tau_l$ is the
751: conformal time when $l=aH$. Since we require this function to be
752: continuous, there must be a matching relation that
753: $u_1(k,\tau_i)=[u_{1max}(k)\cos\phi_l]/a(\tau_l)$. Based on these
754: relations one can get the first factor $F_1$ as follows
755: \begin{eqnarray}\label{F1L}
756: F_1=\left(\frac{1+z(\tau)}{1+z_l}\right)^2\cos^2[l(\tau-\tau_l)+\phi_l]/\cos^2\phi_l,
757: \end{eqnarray}
758: where we introduce the redshift $1+z=a_0/a(\tau)$ in aim of showing
759: the effects of the suppression as a result of redshift. The index
760: ``$0$" indicates today, and $z_l$ is the redshift when the modes
761: re-entered the horizon $l=aH$. Note that the relation of $z_l$ and
762: $k$ can be given by the following equation
763: \begin{eqnarray}
764: \left(\frac{k}{k_0}\right)^2=\sum_i\Omega_i^{(0)}(1+z_1)\exp\left[3\int_0^{z_1}\frac{w_i(\tilde
765: z)}{1+\tilde z}d\tilde z \right],
766: \end{eqnarray}
767: where the sum over $i$ includes all components in the universe.
768: Since the contributions from dark energy and the fluctuations in
769: radiation are very small, here we ignore them and then solve out
770: \begin{eqnarray}\label{zl}
771: 1+z_l=\frac{1+z_{eq}}{2}\left[-1+\sqrt{1+\frac{4(l/k_0)^2}{(1+z_{eq})\Omega_m^{(0)}}}\right],
772: \end{eqnarray}
773: where $z_{eq}\equiv-1+\Omega_m^{(0)}/\Omega_r^{(0)}$.
774: %By the same
775: %process, one can have the right-handed $F_1^R$ that
776: %\begin{eqnarray}\label{F1R}
777: %F_1^R=\left(\frac{1+z(\tau)}{1+z_r(k)}\right)^2\cos^2[l_2(\tau-\tau_r)+\phi_r]/\cos^2\phi_r
778: %\end{eqnarray}
779: %where $\tau_r$ is the conformal time when $l_2=aH$, and $z_r(k)$
780: %satisfy the same relation as Eq. (\ref{zl}) but merely let $l_2$
781: %replaced with $l_1$. Apparently, the factors $F_1^L$ and $F_1^R$
782: The factor $F_1$ describes the redshift-suppressing effect on the
783: primordial gravitational waves. Since this factor shows strongly
784: oscillating behaviour which is inconspicuous to be observed in the
785: GWB, we usually average the term $\cos^2[l(\tau-\tau_l)+\phi_l]$
786: and instead it with $\frac{1}{2}$.
787:
788: Secondly, when considering the influence of the background state
789: of universe on the re-entry of horizon, we focus on analyzing the
790: factor $F_2$. Since the background equation of state $w$ varies
791: very slowly, it is profitable to assume that the evolution of the
792: scale factor is of form $a=a_0(\frac{\tau}{\tau_0})^{\alpha}$ with
793: $\alpha=\frac{2}{1+3w}$. Then solving Eq. (\ref{h equation}) again
794: and ignoring $\Sigma_{ij}$, we have $\tilde
795: u_1(k,\tau)=u_1(k,\tau_i)\Gamma(\alpha+\frac{1}{2})\left(-\frac{l\tau}{2}\right)^{\frac{1}{2}-\alpha}J_{\alpha-\frac{1}{2}}(-l\tau)$,
796: where $\Gamma$ is the Gamma function and $J_{\nu}$ is the $\nu$th
797: Bessel function. If $|l\tau|\gg1$, there is such a relation that
798: $|\frac{\tilde
799: u_1(k,\tau)}{u_1(k,\tau_i)}|^2=\frac{\Gamma^2(\alpha+\frac{1}{2})}{\pi}(-\frac{l\tau}{2})^{-2\alpha}\cos^2(l\tau+\frac{\alpha\pi}{2})$.
800: To match with Eq. (\ref{F1L}), considering that the phase should
801: be continuous, hence we have the solution that when GWB re-enters
802: the horizon the conformal time $\tau_l=-\frac{\alpha}{l}$.
803: Therefore, the second factor $F_2$ is given by
804: \begin{eqnarray}\label{F2L}
805: F_2=\frac{\Gamma^2(\alpha+\frac{1}{2})}{\pi}\left(\frac{2}{\alpha}\right)^{2\alpha}\cos^2\phi_l~.
806: \end{eqnarray}
807: %similarly, the second factor for right-handed mode is given by
808: %\begin{eqnarray}\label{F2R}
809: %F_2^R=\frac{\Gamma^2(\alpha+\frac{1}{2})}{\pi}\left(\frac{2}{\alpha}\right)^{2\alpha}\cos^2\phi_r.
810: %\end{eqnarray}
811: The second factor shows that, when the gravitational waves re-enter
812: the horizon, there is a "wall" lying on the horizon which affects
813: the tensor power spectrum.
814:
815: Thirdly, during the evolution of tensor perturbations, the nonzero
816: anisotropic stress tensor $\Sigma_{ij}$ would more or less bring
817: some effects on the GWB. This effect is proposed by Steven
818: Weinberg\cite{Weinberg03}, and usually the primary ingredients are
819: the freely streaming neutrinos which damp the amplitude of the
820: tensor power spectrum. However, this damping effect just make $P_T$
821: times a constant but do not change the behaviour of the GWB's
822: evolving. In our paper, we adopt $F_3=0.80313$.
823:
824: Ultimately, we have discussed three kinds of leading corrections in
825: the transfer function which make contributions in the evolution of
826: the GWB. Using this transfer function, we are able to connect the
827: primordial gravitational waves with what we observe today.
828:
829: \subsection{Analysis of Today's GWB}
830:
831: In this section, we investigate the power spectrum of the current
832: GWB and the corresponding energy spectrum. To substitute eqs.
833: (\ref{F1L}), (\ref{F2L}), and the damping factor $F_3$ into
834: (\ref{PTtoday}), we can give today's tensor power spectrum as
835: follows,
836: \begin{eqnarray}\label{tensor power today}
837: P_T(k,\tau_0)&=&64\pi G\frac{0.80313}{(1+z_l)^2}\frac{\Gamma^2(\alpha+\frac{1}{2})}{2\pi}\left(\frac{2}{\alpha}\right)^{2\alpha}\nonumber\\
838: &\times&\frac{2^{2\nu-3}}{4\pi^2a_i^2}\frac{\Gamma^2(\nu)}{\Gamma^2(\frac{3}{2})}[(1-\varepsilon)a_iH_i]^{2\nu-1}\frac{k^3}{l^{2\nu}},
839: \end{eqnarray}
840: and from Eq. (\ref{approximate tensor energy}), the present energy
841: spectrum is given by
842: $\Omega_{GW}(k,\tau_0)=\frac{1}{12}\frac{k^2}{(a_0H_0)^2}P_T(k,\tau_0)$.
843:
844: \begin{figure}
845: \begin{center}
846: \includegraphics[width=3.7in]{fig3.eps}
847: \end{center}
848: \caption{The black solid line represents the behaviour of energy
849: spectrum of noncommutative tensor perturbations $\Omega_{GW}$, and
850: the red dash line gives the curve of energy spectrum in standard
851: perturbation theory. The most significant difference is that the
852: solid line form a peak in low frequency while the red line keeps
853: raising up. If the NP is small enough, the peak can appear in the
854: area of CMB Pol's observing. Here we let $\alpha_3=10^{-19}$, and
855: the potential of inflation be $V_{inf}\sim M^4$ where
856: $M=5\times10^{15}$Gev.} \label{fig:Omegacom1}
857: \end{figure}
858:
859: When the frequency of GWB is large enough, one can see that today's
860: noncommutative tensor perturbation power spectrum would agree with
861: the standard theory very well. Consequently, there will be very few
862: signals in the high-frequency range. However, if the frequency
863: reaches the value near NP where the corresponding wave length is
864: around the size of the current horizon, the noncommutative term
865: begins to affect the behaviour of the GWB. Again we require the
866: slow-roll parameter $\varepsilon$ to tend forwards to zero, and
867: assume the potential of inflation to be nearly constant of which the
868: scale is $5\times10^{15}$Gev. Then we give the semi-analytical form
869: of the present energy spectrum of the noncommutative tensor
870: perturbations as follows
871: \begin{eqnarray}\label{Omegasemi}
872: \Omega_{GW}(k,\tau_0)h^2&=&2ek^5l^{-3}\left(-1+\sqrt{1+fl^2}\right)^{-2}~,
873: \end{eqnarray}
874: where $e=2.68563\times10^{14}$ and $f=3.10475\times10^{32}$. In
875: order to make a comparison, we give the corresponding energy
876: spectrum without noncommutative term
877: $\Omega_{GW}^{normal}h^2=2ek^2(-1+\sqrt{1+fk^2})^{-2}$. Note that,
878: when the frequency of GWB is near NP, the term after the parameter
879: '$f$' determines the behaviour of the energy spectrum and the term
880: in the brackets $\left(...\right)$ of Eq. (\ref{Omegasemi}) will
881: never vanish even $k$ approaches zero. Due to that, the energy
882: spectrum of noncommutative GWB form a peak in low frequency and
883: then decay rapidly as mentioned in the beginning. To be more
884: specifically, we show this feature in Fig. \ref{fig:Omegacom1} and
885: see that if tuning NP felicitously the peak can be in the
886: detecting range of next generation of CMB experiments(see CMB Pol
887: \cite{TFCR}). In Fig. \ref{fig:Omegacom3} we choose three groups
888: of NP to see the differences among them. We find that, smaller the
889: NP is, more manifest the peak is. That is to say, it is most
890: possible to detect the features of noncommutativity in the GWB
891: with very minor values of NP. This is consistent with the case
892: that when $\alpha_3$ approaches $0$, the commutative one is
893: recovered. In fact we select $\alpha_3=10^{-19}$ in Fig.
894: \ref{fig:Omegacom1} for the same consideration.
895:
896: \begin{figure}
897: \begin{center}
898: \includegraphics[width=3.7in]{fig4.eps}
899: \end{center}
900: \caption{The black solid line represents the energy spectrum of
901: noncommutative tensor perturbations with the smallest NP; the red
902: dash line gives the curve of energy spectrum with the bigger NP;
903: and the blue dot curve shows the energy spectrum with the biggest
904: NP. One can see that the peak of GWB in black solid line is the
905: most manifest. Here the potential of inflation is taken to be
906: $V_{inf}\sim M^4$ in which $M=5\times10^{15}$Gev.}
907: \label{fig:Omegacom3}
908: \end{figure}
909:
910: \section{Conclusions and Discussions}
911:
912: As a conclusion, we have investigated the key features of the
913: primordial tensor perturbations in the noncommutative field
914: approach, and discussed the possibility of detecting the
915: corresponding GWB in the experiments. Due to the noncommutative
916: effects, the dispersion relation for the primordial gravitational
917: waves is modified and the solution of the the tensor perturbations
918: is different from the commutative case. Therefore, it brings about a
919: lot of exciting phenomena which is brand-new and valuable for us to
920: investigate. To study these new features, we investigated the
921: transfer function to obtain the spectrum of noncommutative
922: gravitational waves that we are observing today. Since the
923: noncommutative term would bring CPT violation and produce effective
924: mass for the graviton, it is reasonable to require this term to be
925: small enough. In our note, this has already been discussed that it
926: is allowed to set the values of NP lower than $10^{-16}$ due to the
927: requirements of both experiments and theories. From the calculations
928: in this note, one can see that one most intriguing effect of
929: noncommutative GWB is that it would generate a peak on its energy
930: spectrum where the frequency may be lower than $10^{-16}$. As a
931: result, on one hand, this phenomenon provides a much more stronger
932: limit on the graviton mass, since we can check the position of this
933: possible peak in the energy spectrum; on the other hand, we expect
934: that the signals of noncommutativity can be found in the next
935: generation of CMB observations if the noncommutativity in the relic
936: tensor perturbations is hidden in the range near current horizon.
937: Eventually, the noncommutativity of gravitational waves definitely
938: go beyond the knowledge of Einstein's gravity and therefore should
939: be an important subject for us to investigate.
940:
941:
942: \acknowledgments
943:
944: It is a pleasure to thank Xinmin Zhang and Bin Chen for helpful
945: discussions, and we thank Yi Wang, Gong-Bo Zhao, and Xiao-Fei
946: Zhang for part of computer calculations. This work is supported in
947: part by National Natural Science Foundation of China under Grant
948: Nos. 90303004, 10533010, 19925523 and 10405029, as well as in part
949: by the Scientific Research Fund of GUCAS(NO.055101BM03).
950:
951: \vfill
952:
953: \begin{thebibliography}{99}
954:
955: \bibitem{Guth81} A. H. Guth, Phys. Rev. {\bf D23}, 347 (1981).
956:
957: \bibitem{Steinhardt82} A. Albrecht, and P. Steinhardt, Phys. Rev.
958: Lett. {\bf 48}, 1220 (1982).
959:
960: \bibitem{Linde82} A. D. Linde, Phys. Lett. {\bf B108}, 389 (1982).
961:
962: \bibitem{CMBobserve} A. D. Miller {\it et al.}, Astrophys. J. {\bf
963: 524}, L1 (1999), astro-ph/9906421; P. de Bernardis {\it et al.},
964: Nature {\bf 404}, 955 (2000), astro-ph/0004404; S. Hanany {\it et
965: al.}, Astrophys. J. {\bf 524}, L5 (2000), astro-ph/0005123; N. W.
966: Halverson {\it et al.}, Astrophys. J. {\bf 568}, 38 (2002),
967: astro-ph/0104489; B. S. Mason {\it et al.}, Astrophys. J. {\bf
968: 591}, 540 (2003), astro-ph/0205384; A. Benoit {\it et al.}, Astro.
969: Astrophys. {\bf 399}, L25 (2003), astro-ph/0210306; J. H.
970: Goldstein {\it et al.}, Astrophys. J. {\bf 599}, 773 (2003),
971: astro-ph/0212517;
972:
973: \bibitem{WMAP} D. N. Spergel {\it et al.}, Astrophys. J. Suppl. {\bf 148}, 175
974: (2003), astro-ph/0302209; D. N. Spergel {\it et al.},
975: astro-ph/0603449.
976:
977: \bibitem{Mukhanov81} V. Mukhanov, and G. Chibisov, JETP {\bf 33}, 549 (1981).
978:
979: \bibitem{Guth82} A. H. Guth, and S.-Y. Pi, Phys. Rev. Lett. {\bf 49}, 1110
980: (1982).
981:
982: \bibitem{Hawking82} S. W. Hawking, Phys. Lett. {\bf B115}, 295 (1982).
983:
984: \bibitem{Starobinsky82} A. A. Starobinsky, Phys. Lett. {\bf B117}, 175 (1982).
985:
986: \bibitem{Bardeen83} J. M. Bardeen, P. J. Steinhardt, and M. S. Turner,
987: Phys. Rev. {\bf D28}, 679 (1983).
988:
989: \bibitem{BBO} URL: http://universe.nasa.gov/program/vision/bbo.html.
990:
991: %\bibitem{DECIGO} N. Seto, S. Kawamura, and T. Nakamura, Phys. Rev.
992: %Lett. {\bf 87}, 221103 (2001), astro-ph/0108011.
993:
994: \bibitem{CMBnextobserve} URL: http://www.planck.fr/.
995:
996: \bibitem{VPJ} L. Verde, H. Peiris, and R. Jimenez,
997: JCAP {\bf 0601}, 019 (2006), astro-ph/0506036.
998:
999: \bibitem{Boyle06} L. Boyle, P. Steinhardt, and N. Turok, Phys. Rev. Lett. {\bf 96}, 111301
1000: (2006), astro-ph/0507455.
1001:
1002: \bibitem{Smith06} T. Smith, M. Kamionkowski, and A. Cooray, Phys. Rev. {\bf D73}, 023504
1003: (2006), astro-ph/0506422.
1004:
1005: \bibitem{Grishchuk75} L. P. Grishchuk, Sov. Phys. JETP {\bf 40},
1006: 409 (1975).
1007:
1008: \bibitem{Allen88} B. Allen, Phys. Rev. {\bf D37}, 2078 (1988).
1009:
1010: \bibitem{Starobinski79} A. Starobinsky, JETP Lett. {\bf 30}, 682
1011: (1979); V. Rubakov, M. Sazhin, and A. Veryaskin, Phys. Lett. {\bf
1012: B115}, 189 (1982); R. Fabbri, and M. Pollock, Phys. Lett. {\bf
1013: B125}, 445 (1983); L. Abbott, and M. Wise, Nucl. Phys. {\bf B244},
1014: 541 (1984); L. Abbott, and D. Harari, Nucl. Phys. {\bf B264}, 487
1015: (1986).
1016:
1017: \bibitem{Lyth93} E. Stewart, and D. Lyth, Phys. Lett. {\bf B302}, 171
1018: (1993), gr-qc/9302019.
1019:
1020: \bibitem{Gasperini93} M. Gasperini, and M. Giovannini, Phys. Rev. {\bf
1021: D47}, 1519 (1993), gr-qc/9211021; R. Brustein, M. Gasperini, M.
1022: Giovannini, and G. Veneziano, Phys. Lett. {\bf B361}, 45 (1995),
1023: hep-th/9507017; A. Buonanno, M. Maggiore, and C. Ungarelli, Phys.
1024: Rev. {\bf D55}, 3330 (1997), gr-qc/9605072.
1025:
1026: \bibitem{Boyle04} L. Boyle, P. Steinhardt, and N. Turok, Phys. Rev.
1027: {\bf D69}, 127302 (2004), hep-th/0307170.
1028:
1029: \bibitem{piao06} Y.-S. Piao, Phys. Rev. {\bf D73}, 047302 (2006),
1030: gr-qc/0601115.
1031:
1032: \bibitem{piao03} Y.-S. Piao, and E. Zhou, Phys. Rev. {\bf D68}, 083515
1033: (2003), hep-th/0308080; Y.-S. Piao, and Y.-Z. Zhang, Phys. Rev.
1034: {\bf D70}, 063513 (2004), astro-ph/0401231.
1035:
1036: \bibitem{pin-others} P. F. Gonzalez-Diaz, and J. A. Jimenez-Madrid, Phys. Lett. {\bf B596}, 16 (2004),
1037: hep-th/0406261; M. Baldi, F. Finelli, and S. Matarrese, Phys. Rev.
1038: {\bf D72}, 083504 (2005), astro-ph/0505552.
1039:
1040: %\bibitem{Bennett} C. L. Bennett {\it et al.}, Astrophys. J. Suppl.
1041: %{\bf 148}, 1 (2003); M. Tegmark {\it et al.}, Astrophys. J. {\bf
1042: %606}, 702 (2004); A. Riess {\it et al.}, Astrophysc. J. {\bf 607},
1043: %665 (2001); M. Tegmark {\it et al.}, Phys. Rev. {\bf D69}, 103501
1044: %(2004), astro-ph/0310723.
1045:
1046: \bibitem{Douglas97} A. Connes, M. Douglas, and A. Schwarz, JHEP {\bf
1047: 9802}, 003 (1998), hep-th/9711162.
1048:
1049: \bibitem{Seiberg99} N. Seiberg, and E. Witten, JHEP {\bf 9909}, 032 (1999),
1050: hep-th/9908142.
1051:
1052: \bibitem{Bigatti00} D. Bigatti, and L. Susskind, Phys. Rev.
1053: {\bf D62}, 066004 (2000), hep-th/9908056; N. Seiberg, L. Susskind,
1054: and N. Toumbas, JHEP {\bf 0006}, 021 (2000), hep-th/0005040.
1055:
1056: \bibitem{Alekseev00} A. Yu. Alekseev, A. Recknagel, and V. Schomerus, JHEP {\bf
1057: 0005}, 010 (2000), hep-th/0003187.
1058:
1059: \bibitem{Chu02} C.-S. Chu, and P.-M. Ho, Nucl. Phys. {\bf B636}, 141 (2002),
1060: hep-th/0203186.
1061:
1062: \bibitem{Seiberg00} N. Seiberg, L. Susskind, and N. Toumbas, JHEP
1063: {\bf 0006}, 044 (2000), hep-th/0005015.
1064:
1065: \bibitem{Gomis00} J. Gomis, and T. Mehen, Nucl. Phys. {\bf B591}, 265 (2000),
1066: hep-th/0005129.
1067:
1068: \bibitem{Carroll01} S. M. Carroll, J. Harvey, V. Kostelecky, C. Lane, and T.
1069: Okamoto, Phys. Rev. Lett. {\bf 87}, 141601 (2001), hep-th/0105082.
1070:
1071: \bibitem{Carson02} C. Carlson, C. Carone, and R. Lebed, Phys. Lett. {\bf B549},
1072: 337 (2002), hep-ph/0209077.
1073:
1074: \bibitem{Chaichian04} M. Chaichian, P. Kulish, K. Nishijima, and A. Tureanu, Phys.
1075: Lett. {\bf B604}, 98 (2004), hep-th/0408069.
1076:
1077: %\bibitem{scalespectrum} Y.-S. Piao, Phys. Lett. {\bf B606}, 245
1078: %(2005), hep-th/0404002; L. E. Allen, and D. Wands, Phys. Rev. {\bf
1079: %D70}, 063515 (2004), astro-ph/0404441; S. Gratton, J. Khoury, P. J.
1080: %Steinhardt, and N. Turok, Phys. Rev. {\bf D69}, 103505 (2004),
1081: %astro-ph/0301395; L. A. Boyle, P. J. Steinhardt, and N. Turok, Phys.
1082: %Rev. {\bf D70}, 023504 (2004), hep-th/0403026; V. Bozza, JCAP {\bf
1083: %0602}, 009 (2006), hep-th/0512066; J. K. Erickson, S. Gratton, P. J.
1084: %Steinhardt, and N. Turok, hep-th/0607164; A. Nayeri, R. H.
1085: %Brandenberger, and C. Vafa, Phys. Rev. Lett {\bf 97}, 021302 (2006),
1086: %hep-th/0511140.
1087:
1088: %\bibitem{othermetric} C. Caprini, and R. Durrer, Phys. Rev. {\bf D65}, 023517 (2002),
1089: %astro-ph/0106244; A. Cooray, P. Corasaniti, T. Giannantonio, and A.
1090: %Melchiorri, Phys. Rev. {\bf D72}, 023514 (2005), astro-ph/0504290;
1091: %T. J. Battefeld, S. P. Patil, and R. H. Brandenberger, Phys. Rev.
1092: %{\bf D73}, 086002 (2006), hep-th/0509043; K. N. Ananda, C. Clarkson,
1093: %and D. Wands, gr-qc/0612013; Y. Watanabe, and E. Komatsu, Phys. Rev.
1094: %{\bf D73}, 123515 (2006), astro-ph/0604176.
1095:
1096: \bibitem{Brandenberger00} R. Brandenberger, and P.-M. Ho, Phys. Rev. {\bf D66}, 023517
1097: (2002), hep-th/0203119; S. Tsujikawa, R. Maartens, and R.
1098: Brandenberger, Phys. Lett. {\bf B574}, 141 (2003),
1099: astro-ph/0308169; G. Calcagni, Phys. Rev. {\bf D70}, 103525
1100: (2004), hep-th/0406006.
1101:
1102: \bibitem{HuangLi} Q. Huang, and M. Li, JHEP {\bf 0306}, 014 (2003),
1103: hep-th/0304203; Q. Huang, and M. Li, JCAP {\bf 0311}, 001 (2003),
1104: astro-ph/0308458; Q. Huang, and M. Li, Nucl. Phys. {\bf B713}, 219
1105: (2005), astro-ph/0311378.
1106:
1107: \bibitem{Cai04} R.-G. Cai, Phys. Lett. {\bf B593}, 1 (2004), hep-th/0403134.
1108:
1109: \bibitem{HuangZhang} Q.-G. Huang, Phys. Rev. {\bf D74} 063513 (2006), astro-ph/0605442;
1110: X. Zhang, JCAP {\bf 0612}, 002 (2006), hep-th/0608207.
1111:
1112: \bibitem{Fatollahi} A. H. Fatollahi, and M. Hajirahimi, Phys. Lett. {\bf B641}, 381
1113: (2006), hep-th/0611225.
1114:
1115: \bibitem{BY} K. Bamba, and J. Yokoyama, Phys. Rev.
1116: {\bf D70}, 083508 (2004), hep-ph/0409237.
1117:
1118: %\bibitem{Violatedec} N. Arkani-Hamed, H. Cheng, M. A. Luty, and S.
1119: %Mukohyama, JHEP {\bf 0405}, 074 (2004), hep-th/0312099; N.
1120: %Arkani-Hamed, P. Creminelli, S. Mukohyama, and M. Zaldarriaga, JCAP
1121: %{\bf 0404}, 001 (2004), hep-th/0312100; R. Buniy, and S. Hsu, Phys.
1122: %Lett. {\bf B632}, 543 (2006), hep-th/0502203.
1123:
1124: \bibitem{Carmona02} J. M. Carmona, J. L. Cortes, J. Gamboa, and F.
1125: Mendez, Phys. Lett. {\bf B565}, 222 (2003), hep-th/0207158.
1126:
1127: \bibitem{Gamboa05} J. Gamboa, and J. Lopez-Sarrion, Phys. Rev. {\bf D71}, 067702 (2005), hep-th/0501034; H.
1128: Falomir, J. Gamboa, J. Lopez-Sarrion, F. Mendez, and A. J. da
1129: Silva, Phys. Lett. {\bf B632}, 740 (2006), hep-th/0504032.
1130:
1131: \bibitem{Nascimento06} J. R. Nascimento, A. Petrov, and R. F. Ribeiro,
1132: hep-th/0601077.
1133:
1134: \bibitem{Arias06} P. Arias, A. Das, J. Gamboa, J. Lopez-Sarrion, and F. Mendez,
1135: hep-th/0610152.
1136:
1137: \bibitem{Ferrari06} A. F. Ferrari, M. Gomes, J. R. Nascimento, E. Passos, A. Yu.
1138: Petrov, and A. J. da Silva, hep-th/0609222.
1139:
1140: \bibitem{Feng06} B. Feng, M. Li, J. Xia, X. Chen, and X. M. Zhang,
1141: Phys. Rev. Lett. {\bf 96}, 221302 (2006), astro-ph/0601095.
1142:
1143: %\bibitem{Mingzhe06} M. Li, J. Xia, H. Li, and X. M. Zhang,
1144: %hep-ph/0611192.
1145:
1146: \bibitem{Weinberg03} S. Weinberg, Phys. Rev. {\bf D69}, 023503
1147: (2004), astro-ph/0306304.
1148:
1149: \bibitem{Pritchard04} J. R. Pritchard, and M. Kamionkowski, Annals Phys. {\bf 318}, 2
1150: (2005), astro-ph/0412581.
1151:
1152: \bibitem{Bashinsky} S. Bachinsky, astro-ph/0505502.
1153:
1154: \bibitem{Dicus05} D. A. Dicus, and W. Repko, Phys . Rev. {\bf D72}, 088302
1155: (2005), astro-ph/0509096.
1156:
1157: \bibitem{Brandenberger92} V. F. Mukhanov, H. A. Feldman, and R. H.
1158: Brandenberger, Phys. Rept. {\bf 215}, 203 (1992).
1159:
1160: \bibitem{BrandenbergerM00} J. Martin, and R. H. Brandenberger, Phys. Rev. {\bf D63}, 123501
1161: (2001), hep-th/0005209; J. Martin, and R. Brandenberger, Phys.
1162: Rev. {\bf D68}, 063513 (2003), hep-th/0305161.
1163:
1164: \bibitem{Gruzinov01} A. Gruzinov, New Astron. {\bf 10}, 311 (2005),
1165: astro-ph/0112246.
1166:
1167: \bibitem{Cutler02} C. Cutler, W. Hiscock, and S. Larson, Phys. Rev. {\bf D67},
1168: 024015 (2003), gr-qc/0209101.
1169:
1170: \bibitem{Cooray03} A. Cooray, and N. Seto, Phys. Rev. {\bf D69}, 103502 (2004),
1171: astro-ph/0311054.
1172:
1173: \bibitem{Jones04} D. I. Jones, Astrophys. J. {\bf 618}, L115 (2004),
1174: gr-qc/0411123.
1175:
1176: \bibitem{Steinhardt05} L. A. Boyle, and P. J. Steinhardt, astro-ph/0512014.
1177:
1178: \bibitem{TFCR} R. Weiss {\it et al.}, Task Force On Cosmic Microwave
1179: Research, (www.science.doe.gov/hep/TFCRreport.pdf).
1180:
1181: %\bibitem{Copeland06} V. Mukhanov, Phys. Rept. {\bf 215}, 203
1182: %(1992); N. Deruelle, and V. F. Mukhanov, Phys. Rev. {\bf D52}, 5549
1183: %(1995), gr-qc/9503050; R. Durrer, and F. Vernizzi, Phys. Rev. {\bf
1184: %D66}, 083503 (2002), hep-ph/0203275; F. Nitti, M. Porrati, and J.-W.
1185: %Rombouts, Phys. Rev. {\bf D72}, 063503 (2005), hep-th/0503247; E. J.
1186: %Copeland, and D. Wands, hep-th/0609183.
1187:
1188: \end{thebibliography}
1189:
1190: \end{document}
1191: