вход по аккаунту


Doubly special relativity in de Sitter spacetime.

код для вставкиСкачать
Ann. Phys. (Berlin) 522, No. 12, 924 – 940 (2010) / DOI 10.1002/andp.201000105
Doubly special relativity in de Sitter spacetime
S. Mignemi1,2,∗
Dipartimento di Matematica, Università di Cagliari, viale Merello 92, 09123 Cagliari, Italy
INFN, Sezione di Cagliari, Cittadella Universitaria, 09042 Monserrato (Cagliari), Italy
Received 10 August 2010, revised 13 September 2010, accepted 28 September 2010 by F. W. Hehl
Published online 26 October 2010
Key words Doubly special relativity, de Sitter space, noncommutative spacetime, de Sitter group.
We discuss the generalization of Doubly Special Relativity to a curved de Sitter background. The model has
three fundamental observer-independent scales, the velocity of light c, the de Sitter radius α, and the Planck
energy κ, and can be realized through a nonlinear action of the de Sitter group on a noncommutative position
space. We consider different choices of coordinates on the de Sitter hyperboloid that, although equivalent,
may be more suitable for treating different problems.
Also the momentum space can be described as a hyperboloid embedded in a five-dimensional space, but
in this case different choices of coordinates lead to inequivalent models. We investigate the kinematics and
the Hamiltonian dynamics of some specific models and describe some of their phenomenological consequences. Finally, we show that it is possible to construct a model exhibiting a duality for the interchange of
positions and momenta together with the interchange of α and κ.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 Introduction
Since the early years of general relativity, de Sitter and anti-de Sitter spacetimes have acquired a fundamental importance, both theoretical and phenomenological, especially in the context of cosmology. Indeed,
recent astrophysical observations seem to indicate that our universe has positive cosmological constant [1].
In spite of their relevance, there is not much literature about the extension of the kinematics of special
relativity to de Sitter or anti-de Sitter backgrounds1. Geometrically, de Sitter space is defined as a space
of constant positive curvature. Its isometries are generated by the de√Sitter algebra, that can be considered
as a deformation of the Poincaré algebra with a parameter α = 1/ Λ with the dimension of length2. Of
course, several geometric and algebraic properties of de Sitter space differ from those of Minkowski space.
For example, contrary to flat space, in de Sitter space the generators of the translations cannot be identified
with the canonical momenta, as should be obvious because of the position dependence of the de Sitter
Hamiltonian. Moreover, there is no natural parametrization of the space and, depending on the specific
problem one is studying, different systems of coordinates can be more suited for the calculations.
A different kind of deformation of special relativity is given by the more recent proposal of deformed
(or doubly) special relativity (DSR) [6]. This theory is based on the generalization of the standard energymomentum dispersion law of particles P 2 = m2 . The deformation is achieved by modifying the action of
the Lorentz group on momentum space through the introduction of a new observer-independent constant κ,
with the dimensions of energy (usually identified with the Planck energy). In this framework, the transformation law of momenta becomes nonlinear, and that of positions momentum dependent. Special relativity
1 To our knowledge, this is discussed only in [2, 3] for a specific choice of coordinates, and more recently in [4, 5] from a
different point of view.
2 In the following, we denote by Λ one third of the cosmological constant.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
is recovered in the limit κ → ∞. Different choices of the deformed dispersion law correspond to different
DSR models and, even imposing suitable physical constraints, there exist in principle infinite inequivalent
realizations of the DSR axioms.
The physical motivations for the introduction of DSR are given by the possibility of explaining some
anomalies observed in high-energy cosmic ray distribution [7] by means of deformed dispersion relations,
and by the theoretical requirement that the Planck energy, which sets the scale for quantum gravity, be
invariant under the symmetries of spacetime3.
Algebraically, DSR theories can be realized in two equivalent ways, either identifying the generators
of the translations with the phase space momenta and then deforming the Poincaré algebra, as in the κPoincaré algebra approach [10], or by maintaining the canonical form of the Poincaré algebra, but making
it act nonlinearly on momentum space, as in the Magueijo-Smolin (MS) model [11]. The second approach
is especially convenient in the case of the de Sitter algebra, where, as mentioned above, even classically
the generators of the translations do not coincide with the canonical momenta.
From a physical perspective, the geometry of the spacetime on which DSR models act has a fundamental
importance. Unfortunately, however, these models are usually defined only in momentum space and the
spacetime geometry is not fixed uniquely from their postulates. Although it is possible to define DSR in
ordinary spacetime, its most natural realization appears nevertheless to be in terms of noncommutative
geometry, with momentum-dependent metric [12–14], and this is the approach that we adopt in this paper.
The simplest way to construct a DSR model starting from canonical special relativity was suggested
in [15]: one can define the physical momenta as functions of auxiliary variables which transform in the
standard way under Lorentz transformations. The deformed transformation laws and dispersion relations
of the physical momenta then follow from this definition. More recently, it has been shown that also the
definition of a suitable noncommutative position space can be obtained by an analogous procedure [16].
Algebraically, de Sitter space and the DSR momentum space have a very similar structure, both being
realized by imposing a quadratic constraint on the coordinates of a five-dimensional space [17]. However,
their physical interpretation is different: de Sitter space has a natural Riemannian structure, and one can
choose arbitrary coordinates on it; momentum space has no such structure, and different parametrizations
cannot be interpreted as physically equivalent, unless further structure is added. In fact, they lead to inequivalent DSR models with different dispersion relations. Of course, a rigorous discussion of this topic
would require a precise definition of the physical momentum. Unfortunately, to our knowledge, an operational definition of momentum measurements in DSR is still lacking.
The aim of this paper is to extend DSR models to the case of a background de Sitter spacetime. The
first example of a DSR deformation of the de Sitter algebra, limited to the momentum sector of phase
space, was given in [18]. A different realization, based on the formalism of quantum groups and extended
to the full phase space was proposed in [19]. Later, the authors of [20] gave still another realization and
discussed some of its phenomenological consequences. However, all these approaches were purely algebraic, since a metric structure on de Sitter space was not introduced. This lack may lead to ambiguities in
the interpretation of the spacetime structure and in the physical predictions of the models.
After the completion of the present work, a more thorough investigation of DSR in a de Sitter background in the context of the κ-Poincaré approach has appeared [21, 22], mainly focused on the interplay
between Planck-scale and curvature effects. In particular, in [22] the possibility to test the effect of the
deformed Poincaré symmetry on the expansion of the universe by observations of gamma-ray bursts has
been discussed.
In this paper, we shall limit ourselves to consider a classical setting based on a phase space realization
of the theory and on the Hamiltonian dynamics of a free particle. We shall discuss the formal aspects of
the theory, without exploring phenomenological implications, except for the extension of some standard
More precisely, in many DSR models, it is not the energy κ that is left invariant by the deformed transformations, but some
components of the 4-momentum. An extreme example of this fact is given by the Snyder model [8, 9], where Lorentz transformations act in the canonical (linear) way, and only the action of the translation generators is nonlinear.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
predictions of flat-space DSR models enforced by the deformed addition law of momenta. More specific
effects related to the cosmological nature of de Sitter spacetime and to the effective energy dependence of
the cosmological constant are left for future work.
Finally, it is interesting to remark that the deformed de Sitter algebra has two invariant scales, beyond
the speed of light. These are the cosmological constant Λ and the Planck energy κ (or equivalently, the
de Sitter radius α ∼ 1025 m and the Planck length 1/κ ∼ 10−35 m). The two scales differ by 60 orders
of magnitude and are related to the opposite extrema of the range of observable physical phenomena. The
origin of such difference is not explained by the present physical theories. One of the models discussed in
this paper possesses a duality for the interchange of α and κ together with the interchange of positions and
The paper is organized as follows: in Sect. 2 we discuss the de Sitter algebra and different parametrizations of de Sitter space, realized as a hyperboloid embedded in five-dimensional spacetime. In Sect. 3 we
discuss the dynamics of a free particle in de Sitter space. Sections 4 and 5 are devoted to the study of the
generalization of the MS model to de Sitter space, using the results of the previous sections. In Sect. 6, a
different generalization of DSR in de Sitter space is considered, related to the Snyder model. An alternative
realization is given in Sect. 7. In Sect. 8, some physical implications of our results are discussed.
Although we shall not consider this issue in detail, all our result can be straightforwardly extended to
the anti-de Sitter case, by changing the sign of the cosmological constant.
We use the following notations: A, B = 0, . . . , 4; μ, ν = 0, . . . , 3; i, j = 1, . . . , 3. Except when dealing
explicitly with the spacetime metric gμν , we always use lower indices, for example Xμ ≡ ημν X μ , where
X μ are the natural (contravariant) coordinates. The manipulations of indices are always performed with the
flat metric ημν = diag (1, −1, −1, −1), and not with the metric gμν . The product between two 4-vectors
ημν V μ W ν = η μν Vμ Wν is denoted by V ·W , and if V = W by V 2 . For 5-vectors, we write the indices
2 de Sitter space
We review some properties of de Sitter space and its symmetry group, which are not easily found in the
literature. In particular, we discuss some coordinate systems that will be useful in the following. Unfortunately, contrary to Minkowski space, de Sitter space does not admit a natural choice of coordinates. In
particular, as we shall see, different quantities have simpler expressions in different coordinate systems.
We shall therefore alternate between them, depending on the subject under consideration.
2.1 Generalities
= −α2 embedded in
It is well known that de Sitter space can be realized as a hyperboloid of equation ξA
5-dimensional flat space, with coordinates ξA and metric tensor ηAB = diag (1, −1, −1, −1, −1).
The isometries of de Sitter space are generated by the de Sitter algebra. This can be identified with
the Lorentz algebra so(1, 4) of the 5-dimensional space, which leaves invariant the hyperboloid. In terms
of the 5-dimensional canonical positions ξA and momenta πA , the generators of the Lorentz algebra read
JAB = ξA πB − ξB πA , and obey the Poisson brackets
{JAB , JCD } = ηBC JAD − ηBD JAC + ηAD JBC − ηAC JBD .
In mathematical language, de Sitter space can be described as a coset space SO(1, 4)/ SO(1, 3). In
order to parametrize it, one must single out one of the (equivalent) spatial coordinates, say ξ4 . The position
coordinates Xμ of de Sitter space are then obtained through a projection from the hyperboloid to a plane
orthogonal to the ξ4 axis. Moreover, the canonically conjugate momenta Pμ can be defined in such a
way that they obey canonical Poisson brackets with the Xμ , according to the 5-dimensional symplectic
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
Under projection, the original 5-dimensional Lorentz algebra so(1, 4) reduces to the
√ 4-dimensional
Lorentz algebra so(1, 3) generated by the Jμν , while the remaining generators Tμ = Λ J4μ are interpreted as generators of de Sitter translations. The de Sitter algebra can therefore be written as
{Jμν , Jρσ } = ηνσ Jμρ − ηνρ Jμσ + ημρ Jνσ − ημσ Jνρ ,
{Jμν , Tλ } = ημλ Tν − ηνλ Tμ ,
{Tμ , Tν } = −ΛJμν .
Hence, the Lorentz subalgebra of the 4-dimensional de Sitter algebra is identical to the flat space Lorentz
algebra. In particular, if one writes its generators Jμν in terms of the 4-dimensional coordinates Xμ and
their canonically conjugate momenta Pμ , one gets the canonical expression Jμν = Xμ Pν − Xν Pμ . Thus
the positions and the momenta obey the usual transformation laws under Lorentz transformations,
{Jμν , Xλ } = ημλ Xν − ηνλ Xμ ,
{Jμν , Pλ } = ημλ Pν − ηνλ Pμ .
As we shall see, the realization of the translation generators Tμ depends instead on the specific choice
of coordinates on the hyperboloid.
2.2 Natural coordinates
The de Sitter hyperboloid can be parametrized by arbitrary coordinates and, contrary to the case of flat
space, there is no privileged system of coordinates for de Sitter space. The systems commonly used in the
applications to general relativity single out the time coordinate, while the most interesting for our purposes
are isotropic in space and time. Since these systems of coordinates are not very well known, we shortly
review their properties.
We start by considering the natural parametrization, given by X̂μ = ξμ . It follows that Λ ξ4 =
1 + ΛX̂ 2 , and the metric induced on the hyperboloid by the 5-dimensional flat metric reads
ĝμν = ημν −
Λ X̂μ X̂ν
1 + ΛX̂ 2
ĝ μν = η μν + Λ X̂ μ X̂ ν .
A spacelike coordinate singularity, that can be interpreted as a cosmological horizon, occurs at X̂ 2 =
From the definition of Tμ = Λ J4μ , it is easy to see that under translations
{Tμ , X̂ν } = − 1 + ΛX̂ 2 ημν .
The nontrivial effect of translations is of course due to the curvature of the space.
From (2.5) it is evident that the translation generators Tμ do not coincide with the momenta P̂μ = πμ
canonically conjugate to X̂μ . In fact,
Tμ = 1 + ΛX̂ 2 P̂μ ,
and the momenta transform as
ΛX̂ν P̂μ
{Tμ , P̂ν } = .
1 + ΛX̂ 2
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
2.3 Conformal coordinates
The parametrization that yields
√ the simplest form of the metric is given by√conformal (stereographic)
coordinates X̃μ = 2ξμ /(1+ Λ ξ4 ), with inverse ξμ = X̃μ /(1−ΛX̃ 2/4), and Λ ξ4 = (1+ΛX̃ 2/4)/(1−
ΛX̃ 2 /4). In these coordinates the metric takes the diagonal form
ΛX̃ 2
g̃ = 1 −
η μν .
g̃μν =
(1 − Λ X̃ 2 /4)2
The coordinate singularity at X̃ 2 = 4/Λ corresponds to the infinity of the hyperboloid.
Under translations the position coordinates transform as
ΛX̃ 2
ημν + X̃μ X̃ν ,
{Tμ , X̃ν } = − 1 +
and hence, in terms of the canonical momenta P̃μ = (1 + Λ ξ4 ) πμ /2,
ΛX̃ 2
P̃μ − X̃·P̃ X̃μ .
Tμ = 1 +
The momenta transform as
{Tμ , P̃ν } = − (X̃·P̃ ημν + X̃μ P̃ν − X̃ν P̃μ ).
2.4 Beltrami coordinates
Another useful√parametrization of the de Sitter
given by Beltrami
(projective) coordinates [2,
√hyperboloid is√
3], X̄μ = ξμ / Λ ξ4 , with inverse ξμ = X̄μ / 1 − ΛX̄ 2 and Λ ξ4 = 1/ 1 − ΛX̄ 2 . In these coordinates,
the metric has the form
(1 − ΛX̄ 2 )ημν + ΛX̄μ X̄ν
μ ν
ḡμν =
(1 − ΛX̄ 2 )2
Again, the coordinate singularity at X̄ 2 = 1/Λ corresponds to the infinity of the hyperboloid.
Under translations, the coordinates X̄μ transform as
{Tμ , X̄ν } = −ημν + ΛX̄μ X̄ν .
Tμ = P̄μ − Λ X̄·P̄ X̄μ ,
{Tμ , P̄ν } = −Λ(X̄·P̄ ημν + X̄μ P̄ν ).
In terms of the canonical momenta P̄μ = Λ ξ4 πμ , the translation generators read
3 Motion in de Sitter space
We consider now the motion of a free particle in de Sitter space in the coordinate systems introduced in the
previous section, using the Hamiltonian formalism. For simplicity, we do not explicitly impose the Hamiltonian constraint, since it is not essential for our considerations, but this can be done straightforwardly.
We observe also that one may write the dynamics in terms of the conserved quantities Tμ instead of
Pμ . However since the Tμ do not satisfy canonical Poisson brackets with the position coordinates Xμ , the
formalism would become more involved and the generalization to DSR uneasy. Moreover, it is not evident
to us which of the variables should be identified with the physically measured momentum.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
3.1 Generalities4
The lagrangian of a free particle of mass m, invariant under de Sitter transformations, is given by
gμν Ẋ μ Ẋ ν ,
where a dot denotes derivative with respect to an evolution parameter τ .
Varying with respect to Xμ one obtains the geodesics equations. Alternatively, defining the canonically
conjugate momenta
= m gμν Ẋ ν ,
∂ Ẋ μ
Pμ ≡
1 μν
(with inverse Ẋ μ = m
g Pν ), one can define the Hamiltonian as
H = Pμ Ẋ μ − L =
1 μν
g Pμ Pν .
The equations of motion in Hamiltonian form are
Ẋ μ = {X μ , H} = g μν Pν ,
Ṗμ = {Pμ , H} = −
∂g λν
Pλ Pν ,
∂X μ
and can be derived by varying the action
dτ (Ẋ μ Pμ − H).
An equivalent way to obtain the Hamiltonian is to identify it with the quadratic Casimir invariant of
the de Sitter group, JAB JAB = Jμν Jμν − Λ2 Tμ Tμ , written in terms of the four-dimensional phase space
variables. It is easy to see that in this way one recovers the previous results.
3.2 Conformal coordinates
These coordinates give the simplest relation between velocity and momentum. The Hamiltonian takes the
form (from now on we put m = 1),
ΛX̃ 2
P̃ 2 ,
H̃ =
with field equations
X̃˙ μ =
ΛX̃ 2
P̃μ ,
P̃ 2 X̃μ .
P̃˙μ =
In these coordinates, the 3-velocity vi is given by
vi ≡
X̃˙ i P̃i
= ,
X̃˙ 0 P̃0
and hence the relation between 3-velocity and momenta is the same as in flat space.
In this subsection we restore the difference between upper and lower indices.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
The field equations may be integrated by substituting the first Eq. (3.7) into the second. However, it is
more convenient to use the conservation law associated with the translations, which gives a first integral
Tμ = Aμ , with Aμ a constant vector. From (2.10) and (3.7),
(1 + ΛX̃ 2 /4)X̃˙ μ − Λ X̃·X̃˙ X̃μ /2
ΛX̃ 2
P̃μ − X̃·P̃ X̃μ =
= Aμ .
(1 − ΛX̃ 2 /4)2
Inverting, one obtains
1 − ΛX̃ 2 /4
X̃˙ μ =
1 + ΛX̃ 2 /4
ΛX̃ 2
Aμ − A·X̃ X̃μ ,
and therefore
vi =
(1 − ΛX̃ 2 /4)Ai − Λ A·X̃ X̃i /2
(1 − ΛX̃ 2 /4)A0 − Λ A·X̃ X̃0 /2
3.3 Beltrami coordinates
These coordinates have the nice property that 3-dimensional geodesics are straight lines. The Hamiltonian
takes the form
H̄ =
(1 − ΛX̄ 2 )[P̄ 2 − Λ(X̄·P̄ )2 ],
with equations of motion
X̄˙ μ = (1 − ΛX̄ 2 )(P̄μ − Λ X̄·P̄ X̄μ ),
P̄˙μ = Λ P̄ 2 − Λ(X̄·P̄ )2 X̄μ + (1 − ΛX̄ 2 )X̄·P̄ P̄μ ) .
Hence the 3-velocity can be written as
vi ≡
X̄˙ i
P̄i − ΛX̄·P̄ X̄i
X̄0 P̄0 − ΛX̄·P̄ X̄0
Its form no longer coincides with its flat space analogous.
The Eqs. (3.13) are rather involved, but one can exploit the conservation law Ṫμ = 0 to obtain a first
P̄μ − ΛX̄·P̄ X̄μ =
X̄˙ μ
= Aμ ,
1 − ΛX̄ 2
for constant Aμ . Inverting, one obtains
X̄˙ μ = (1 − ΛX̄ 2 )Aμ ,
and hence
vi =
Therefore, free particles have constant 3-velocity and their trajectories in 3-space are straight lines.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
3.4 Natural coordinates
These coordinates do not give rise to particularly simple expressions. Therefore, we just summarize the
main results. The Hamiltonian has the form
Ĥ =
1 2
[P̂ + Λ(X̂·P̂ )2 ],
and yields the equations of motion
X̂μ = P̂μ + Λ X̂·P̂ X̂μ ,
P̂μ = −Λ X̂·P̂ P̂μ ,
with 3-velocity
vi =
P̂i + ΛX̂·P̂ X̂i
P̂0 + ΛX̂·P̂ X̂0
One can again exploit the conservation law Ṫμ = 0 to obtain a first integral,
P̂μ + ΛX̂·P̂ X̂μ =
(1 + ΛX̂ 2 )X̂μ − ΛX̂·X̂ X̂μ
= Aμ .
1 + ΛX̂ 2
Inverting, one obtains X̂μ and then
vi =
Ai − Λ A·X̂ X̂i
A0 − Λ A·X̂ X̂0
4 The MS model in de Sitter space
DSR theories in flat space can be implemented in two different ways. One can either deform the Poincaré
algebra [10, 12], imposing nonlinear Poisson brackets between the generators, or maintain the canonical
form of the algebra, but modify its action on the momentum variables [11, 15]. The first approach has been
considered in [18–20] in order to derive a deformed de Sitter algebra. However, for the discussion of the
extension of DSR models to the full phase space, especially in the case of a de Sitter background, the
second approach appears to be more useful.
The MS model was introduced in [11] and is characterized by a deformed dispersion relation p2 /(1 −
p0 /κ)2 = m2 . A remarkable property of this model is that the Planck energy κ is left invariant under the
deformed Lorentz transformations. The covariant realization of the model in a noncommutative position
space was discussed in [13, 14].
In [16] it was observed that the representation of the MS algebra in phase space can be obtained in a
straightforward way from the Poincaré algebra acting canonically on a space of coordinates Xμ , Pμ , by
performing the substitution
Xμ = (1 − p0 /κ) xμ ,
Pμ =
1 − p0 /κ
pμ =
1 + P0 /κ
with inverse
xμ = (1 + P0 /κ) Xμ ,
Here xμ , pμ are interpreted as physical observables, in contrast with the auxiliary variables Xμ , Pμ .5
Note that this cannot be considered simply as a change of variables. The transformation is noncanonical and the structure of
spacetime is totally different from the classical one.
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
The symplectic structure of phase space is then deformed and takes the form [13, 14],
{xi , xj } = 0,
{p0 , pi } = {pi , pj } = 0,
{x0 , xi } = ,
{xi , pj } = −δij ,
{x0 , p0 } = 1 − ,
{xi , p0 } = 0.
{x0 , pi } = − ,
In particular, the position coordinates xμ do not commute.
One can apply the same procedure to the case of a de Sitter background. In this context it is useful to
rewrite the de Sitter algebra in the form
{Ni , Nj } = ijk Mk ,
{Mi , Nj } = ijk Nk ,
{Ti , Tj } = −Λ ijk Mk ,
{Mi , Tj } = ijk Tk ,
{Ni , Tj } = δij T0 ,
{Mi , Mj } = ijk Mk ,
{T0 , Tj } = −Λ Nj ,
{Mi , T0 } = 0,
{Ni , T0 } = Ti ,
where Mk = 12ijk Jij are the generators of rotations and Ni = J0i the generators of boosts.
The Poisson brackets between phase space variables maintain the form (4.3). Also the deformed action
of the Lorentz subalgebra on coordinates and momenta is the same as in the flat space MS model [14],
{Mi , xj } = ijk xk ,
{Mi , x0 } = 0,
{Ni , xj } = δij x0 + pi xj /κ,
{Ni , x0 } = xi + pi x0 /κ.
{Mi , pj } = ijk pk ,
{Mi , p0 } = 0,
{Ni , pj } = δij p0 − pi pj /κ,
{Ni , p0 } = pi − pi p0 /κ.
The action of translations on coordinates and momenta depends instead on the specific coordinates chosen
for de Sitter space. For example, in the natural parametrization,
Tμ = (1 − p̂0 /κ)−2 + Λx̂2 p̂μ ,
x̂0 x̂ν p̂μ
{Tμ , x̂ν } = − (1 − p̂0 /κ) + Λx̂ ημν −
κ (1 − p̂0 /κ)−2 + Λx̂2
Λ (x̂ν − p̂ν x̂0 /κ) p̂μ
{Tμ , p̂ν } = .
(1 − p̂0 /κ)−2 + Λx̂2
An interesting physical implication of this model is that the cosmological constant becomes effectively
energy dependent. Consider for example natural coordinates and define, in analogy with (4.2), x̂4 = (1 +
P̂0 /κ) X̂4 , with X̂4 = ξ4 . Then x̂2A = −α2 /(1 − p̂0/κ)2 ≡ −1/Λ(p̂0). In particular, for p̂0 → κ, Λ(p̂0 ) →
0, i.e. particles with energy close to the Planck energy do not experience the curvature of spacetime.
5 Dynamics of the MS model in de Sitter space
Also the Hamiltonian of a free particle can be obtained by substituting (4.1) into the undeformed Hamiltonian [16]. The equations of motion can then be obtained by taking into account the deformed symplectic
structure (4.4), namely,
p0 ∂H pi ∂H xi ∂H
ẋ0 = {x0 , H} = 1 −
κ ∂p0 κ ∂pi κ ∂xi
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
ẋi = {xi , H} = −
∂H xi ∂H
∂pi κ ∂x0
p0 ∂H
ṗ0 = {p0 , H} = − 1 −
κ ∂x0
ṗi = {pi , H} =
∂H pi ∂H
∂xi κ ∂x0
Equivalently, the Hamilton equations can be obtained by varying the action in which the substitution (4.1)
has been done.
For example, in conformal coordinates the Hamiltonian is given by
H̃ = Δ̃2 p̃2 ,
Δ̃ =
− (1 − p̃0 /κ) x̃2 .
1 − p̃0 /κ 4
The Hamilton equations then read
x̃˙ μ = Δ̃2 p̃μ + (1 − p̃0 /κ) Δ̃ p̃2 x̃0 x̃μ ,
p̃˙ μ = (1 − p̃0 /κ) Δ̃ p̃2 (x̃μ − x̃0 p̃μ /κ).
They can also be recovered from the action
I = dτ X̃μ P̃μ − (1 − ΛX̃ /4) P̃
1 2 2
(1 − p̃0 /κ) x̃μ − Δ̃ p̃ .
= dτ
1 − p̃0 /κ dτ
The Hamilton equations (5.5) have acquired complicated terms proportional to Λ/κ, and are no longer
linear in the momentum, so that it is not easy to invert them in order to obtain p̃μ in terms of x̃˙ μ . Because
of this, it is difficult to obtain the equations of motion in second order form, even using the conservation
law for Tμ .
Moreover, the property that the velocity has the same expression as in the undeformed case, valid for
the flat space MS model [14], does not extend to the de Sitter case. In fact, this property was proven in [23]
to hold for position-independent Hamiltonians. If one wishes to maintain its validity, one should look for
a different deformation of the symplectic structure. For the same reason, contrary to the flat space MS
model, the evolution parameter dτ cannot be identified with the line element invariant under the deformed
transformations, which reads
ds2 =
(1 − p̃0 /κ)2 dx̃2
2 .
1 − Λ4 (1 − p̃0 /κ)2 x̃2
An analogous calculation can be performed in natural coordinates. The deformed Hamiltonian is
Ĥ =
2 (1 − p̂0 /κ)2
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
with Hamilton equations
x̂˙ μ = (1 − p̂0 /κ)−2 p̂μ + Λ(1 − p̂0 /κ) x̂·p̂ x̂μ ,
p̂˙ μ = −Λ(1 − p̂0 /κ) x̂·p̂ p̂μ .
The invariant metric reads
ds2 = (1 − p̂0 /κ)2 dx̂2 −
(1 − p̂0 /κ)−2 + Λx̂2
Also in this case one finds the same problems as with conformal coordinates.
It is also interesting to notice that the metric (5.11) exhibits a momentum-dependent cosmological horizon at Λx̂2 = −(1 − p̂0 /κ)−2 . This property is of course related to the momentum dependence of the
cosmological constant discussed at the end of previous section. The metrics (5.8) and (5.11) can be considered as examples of rainbow metrics [24].
Finally, we notice that in the limits κ → ∞ and Λ → 0 one recovers the ordinary de Sitter space and
the flat space MS model, respectively, while the limit p0 → κ is analogous to that of the MS model [11].
6 DSR in de Sitter space in a Snyder-like basis
It is known that DSR theories can be realized in several different ways. An interesting realization is given
by the so-called Snyder basis [8], which is characterized by the dispersion relation P 2 /(1 − P 2/κ2 ) = m2 .
In a DSR interpretation [9] this implies that the rest mass of particles must always be less than κ. Another
important property of this basis is that only the action of the translations is deformed, while that of the
Lorentz group is not affected. This example illustrates the fact that the most relevant characteristic for the
implementation of DSR is the deformation of the action of translations (and hence a modified composition
law of momenta) rather than that of Lorentz transformations, as usually postulated.
6.1 Minkowski space
Let us briefly review the case of flat spacetime [9]. It is easy to see that, in analogy with our previous
treatment of the MS model, the easiest way to recover the Snyder realization of DSR is to define new
coordinates Xμ , Pμ starting from the canonical Xμ , Pμ , which are thus interpreted as auxiliary variables,
Xμ =
1 + ΩP 2 Xμ ,
Pμ = √
1 + ΩP 2
where Ω = 1/κ2 is the Planck area6. The inverse transformations are
Xμ = 1 − Ω P 2 Xμ ,
Pμ = √
1 − Ω P2
One has then,
{Xμ , Xν } = −Ω(Xμ Pν − Xν Pμ ),
{Pμ , Pν } = 0,
{Xμ , Pν } = ημν − Ω Pμ Pν . (6.3)
the canonical form (2.3). The translation generaThe Lorentz transformations acting on Xμ , Pμ maintain
tors Tμ must instead be identified with Pμ = Pμ / 1 − Ω P 2 . Their action changes accordingly,
{Tμ , Xν } = √
1 − ΩP 2
{Tμ , Pν } = 0.
In principle one may choose a negative sign for Ω, obtaining an inequivalent model with rather different properties [9]: for
example, in that case the mass does not admit a maximal value, but in the quantum theory a nonzero minimal uncertainty for
the position appears due to the modified commutation relations, as in the model discussed in [25].
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
The invariant Hamiltonian for a free particle can be written as
P2 1
2 1 − Ω P2
with equations of motion
X˙μ =
1 − Ω P2
Ṗμ = 0.
It follows that Ẋμ = Aμ is constant. The 3-velocity is then given by
vi =
Pi Ai
= ,
P0 A0
and the 3-dimensional geodesics are straight lines. Moreover, it is easy to verify that the invariant line
element ds2 = (1 − Ω P 2 )dX 2 can be identified with dτ 2 , with τ the evolution parameter.
6.2 de Sitter space
Let us now extend the above construction to the case of de Sitter space in the Beltrami coordinates of
Sect. 2.4. The substitution (6.1) yields
X̄μ = 1 + ΩP̄ 2 X̄μ = Ω π 2 + (Λξ42 )−1 ξμ ,
P̄μ = √
Ω π + (Λξ42 )−1
1 + ΩP̄
where ξA are as usual the coordinates of the five-dimensional embedding space. The new phase space
coordinates X̄μ , P̄μ satisfy the Poisson brackets (6.3).
Using the definition of ξ4 and inverting (6.8), one gets
ξμ =
πμ = ΦP̄μ ,
(1 − Ω P̄ 2 )−1 − ΛX̄ 2 .
In terms of the variables X̄μ , P̄μ , the Lorentz generators of the de Sitter algebra (2.2) have canonical
form, while the translation generators read
Tμ = √
P̄μ − Λ(1 − Ω P̄ 2 )X¯ ·P̄ X¯μ ,
1 − Ω P̄ 2
{Tμ , X¯ν } = −√
ημν − Λ(1 − Ω P̄ 2 )2 X̄μ X¯ν + ΛΩ(1 − Ω P̄ 2 )X¯ ·P̄ P̄μ X¯ν ) ,
1 − Ω P̄
{Tμ , P̄ν } = −Λ 1 − Ω P̄ 2 X̄ ·P̄(ημν − Ω P̄μ P̄ν ) − (1 − Ω P̄ 2 )X¯μ P̄ν .
One can also define a Hamiltonian, invariant under the full deformed de Sitter group,
1 2 2
Φ P̄ − Λ(1 − Ω P̄ 2 )(X¯ ·P̄)2 .
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
The Hamilton equations take an extremely involved form, and we shall not report them here. The invariant
metric for this model is
gμν =
Φ2 ημν + ΛX¯μ X̄ν
No coordinate singularities occur at finite distance. In the limits Ω → 0 and Λ → 0 one recovers the
ordinary de Sitter space and the flat space Snyder model of previous section, respectively.
7 A different Snyder-like realization
The Snyder realization of DSR in de Sitter space given in the previous section is rather awkward. In this
section, we consider a slightly different realization, which takes a more symmetric form and gives rise to
more elegant formulas. The algebra of this model displays some similarities with that proposed in [20].
As in that case, in the present realization the momenta coincide in the limit Ω → 0 with the de Sitter
translation generators rather than with the canonical momenta, and hence have nontrivial Poisson brackets
with the position coordinates in this limit.
We define
1 + Ωπ 2
1 + Λξ 2
πμ ,
X̆μ =
1 + Λξ 2
1 + Ωπ 2
with inverse
ξμ =
1 − Ω P̆ 2 ˘
Xμ ,
1 − Λ X̆ 2
πμ =
1 − Λ X̆ 2
1 − Ω P̆ 2
P̆μ .
The coordinates (7.1) satisfy the Poisson brackets
{X̆μ , X̆ν } = −
{P̆μ , P̆ν } = −
Ω (1 − Λ X̆ 2 )
1 − ΛΩ X̆ 2 P̆ 2
Λ (1 − Ω P̆ 2 )
1 − ΛΩ X̆ 2 P̆ 2
{X̆μ , P̆ν } = ημν −
(X̆μ P̆ν − X˘ν P̆μ ),
(X̆μ P̆ν − X˘ν P̆μ ),
Λ (1 − Ω P̆ 2 )X˘μ X˘ν + Ω (1 − Λ X̆ 2 )P̆μ P̆ν
1 − ΛΩ X̆ 2 P̆ 2
The Lorentz generators of the de Sitter algebra have canonical form, while the dilatation generators are
1 − ΛΩ X̆ 2 P̆ 2
P̆μ ,
Tμ =
1 − Ω P̆ 2
and their action is given by
2 ˘
2 P̆ 2
X˘μ X̆ν + Ω
{Tμ , X˘ν } = −
ημν −
1 − Ω P̆ 2
1 − ΛΩ X̆ 2 P̆ 2
1 − ΛΩ X̆ 2 P̆ 2
2 ˘
1 − Ω P̆ 2
X̆μ P̆ν − X˘ν P̆μ + Ω
{Tμ , P̆ν } = −Λ
1 − ΛΩ X̆ 2 P̆ 2
1 − ΛΩ X̆ 2 P̆ 2
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
The Hamiltonian of a free particle can be obtained from the Casimir invariant of the de Sitter algebra,
and takes the form
1 1 − ΛX̆ 2 2
P̆ + Λ(X̆ ·P̆)2 .
2 1 − Ω P̆ 2
Taking into account the symplectic structure (7.3), the equations of motion ensuing from the Hamiltonian
ΛΩ X̆ ·P̆ P̆ 2 X˘μ
X̆μ = (1 − Λ X̆ )
1 − Ω P̆ 2 1 − ΛΩ X̆ 2 P̆ 2
ΛΩ(1 − Λ X̆ ) ˘
X ·P̆ P̆ 2 P̆μ .
P̆μ =
1 − ΛΩ X̆ 2 P̆ 2
Also in this case, there does not seem to exist a simple relation between velocity and momentum.
The 4-dimensional metric can be derived in the usual way from the 5-dimensional flat metric subject to
the constraint
1 − ΛΩ X̆ 2 P̆ 2
ξ4 = 1 + Λξ 2 =
1 − Λ X̆ 2
and reads
gμν =
1 − Ω P̆ 2
1 − Λ X̆ 2
ημν + Λ
1 + Ω(1 − Λ X̆ 2 )P̆ 2
(1 − Λ X̆ 2 )(1 − ΛΩ X̆ 2 P̆ 2 )
Xμ X̆ν .
Also in this case there is no evident relation between the metric and the differential dτ of the evolution
parameter. From (7.2) it is evident that the coordinate singularity at X̆ 2 = 1/Λ corresponds to points at
infinity in the hyperboloid. Moreover, the metric (7.9) presents a second momentum-dependent coordinate
singularity at ΛX̆ 2 = 1/ΩP̆ 2, or better X̆ 2 P̆ 2 = 1/ΛΩ. However, for such values of X˘ and P̆ the model
is ill-defined (see (7.3)): this region is anyway far beyond the range of physically observable phenomena,
since 1/ΛΩ ∼ 10120 .
In the limit Λ → 0 one of course recovers the flat-space Snyder model of previous section, while in the
limit Ω → 0 one gets the standard de Sitter space, although with noncanonical Poisson brackets between
positions and momenta (since the momenta
are identified
generators in this limit).
√ with the translation
More interesting are the limits X̆ → 1/ Λ and P̆ → 1/ Ω. For X˘ → 1/ Λ, one is close to infinity, √
the symplectic structure reduces to the undeformed one obtained in the limit Ω = 0. The limit P̆ → 1/ Ω
corresponds instead to the extremal value of the momentum. In this limit, the symplectic structure is that
of the flat Snyder model, Λ = 0, and the metric and the Hamiltonian become singular.
It is also interesting to notice that the Poisson brackets (7.3) lead after quantization to generalized
commutation relations of the most general kind proposed in [25] that, in case of negative Λ and Ω, imply
the existence of both a minimal length and momentum.
Another interesting property of this model is the existence of a duality for the exchange of X̆ ↔ P̆,
together with Λ ↔ Ω. This duality connects the high-energy/short-distance regime, governed by the Planck
area Ω, with the low-energy/long-distance regime, governed by the cosmological constant Λ.
8 Phenomenological implications
We briefly review some immediate phenomenological consequences of our model, analogous to those of
flat space DSR. We also mention some effects that could be associated to the cosmological character of
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
de Sitter spacetime, but a more thorough investigation of this topic is left for future work. The discussion
relies on our specific interpretation of the DSR formalism, that, in particular, implies a constant speed of
light. Other interpretations may lead to different predictions (see for example [26]).
8.1 Addition law of momenta
One of the main phenomenological predictions of DSR is a deformation of the addition law of momenta.
For consistency with the DSR postulates, in fact, the total energy or momentum of a composite system
should not exceed κ. This fact poses the problem of the transition to macroscopic systems, where such
limit cannot be valid, but this is beyond the scope of the theory, at least if it is considered just as a phenomenological model describing quantum gravity effects.
The main application of the deformed addition law is to the scattering of ultra-high energy particles by
the background cosmological radiation, which may give rise to observable effects [7].
To obtain the addition law, one writes the conserved quantity Tμi for the scattered particles as a function
Tμ . The total momentum ptμ is
of their physical momenta piμ as Tμi = Tμi (piμ ), and then adds, Tμt =
finally obtained by inverting the previous relation and writing pμ = pμ (Tμ ) [15, 25].
In de Sitter relativity, Tμ (pμ ) is a (position-dependent) linear function of pμ and therefore the physical
momenta add in the usual linear way, at least if x is fixed, as in the case of point interactions (in any case
corrections could be relevant only at cosmological scales).
In de Sitter DSR the function Tμ (pμ ) is nonlinear and position dependent. However, for point interactions one can always choose the origin as the collision point. In this case, the addition law reduces to that
of the corresponding flat space DSR model.
For example, in the MS model of Sects. 4–5, one obtains the addition law [15]
p /(1 − pi0 )
μ i
ptμ =
1 + p0 /(1 − pi0 )
Its physical consequences have been thoroughly investigated in [27].
If the scattering point is not at the origin, Tμ (pμ ) is given by (4.6), and its inversion is nontrivial, so that
an explicit expression for the addition law is awkward. Due to the homogeneity of de Sitter space, one may
however always perform a translation in order to carry x into the origin.
8.2 Proper time
The possibility of defining an invariant metric allows one to easily calculate the correction to the lifetime
of a massive particle. In particular, in conformal coordinates the metric has the simple expression
ds2 = A2 (x, p) dx2 ,
and, calling xμ , pμ the laboratory coordinates and xμ , pμ the rest coordinates of the particle, one has
A2 (x , p )dt2 = A2 (x, p)(1 − v2 )dt2 ,
where we have put x0 = t. Hence,
dt =
A(x , p )
γ dt
A(x, p)
with γ = (1 − v2 )−1/2 . This gives rise to corrections to the relativistic formula. For example, for the model
of Sects. 4–5, A(x, p) = 1/Δ̃, (cf. (5.8)) and
(1 − M/κ) dt
(1 − p0 /κ) dt
γ 1 − 4γΛ2 (1 − p0 /κ)2 t2
1 − Λ4(1 − M κ)2 t2
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Ann. Phys. (Berlin) 522, No. 12 (2010)
where M ≡ p0 is the rest energy of the particle (not to be confused with its Casimir mass m). Integrating,
one gets the same relation that holds in the flat case [28],
(1 − M/κ)
p0 − M
Δt =
γ Δt ∼ 1 +
γ Δt .
(1 − p0 /κ)
One has therefore energy-dependent corrections to the formula relating proper time to laboratory time, and
hence to the observed lifetime of high-energy particles. The corrections are however of order p0 /κ and are
only relevant for Planck-scale energies.
8.3 Cosmology
As mentioned before, in de Sitter DSR the cosmological constant becomes effectively energy dependent.
This can have important consequences on the cosmological models, especially in the first stages of the
cosmological evolution, when Planck-scale energies are involved.
Moreover, also astrophysical observations may be affected, since ultra-high energy particles experience
a smaller value of the cosmological constant. Also the cosmological horizon is shifted, depending on the
energy of the particles. This may have implications on the horizon problem, as observed in [11].
Recently, the interplay between curvature effects and DSR has been investigated in great detail in [21,
22] in the context of the κ-Poincaré formalism. It must be noticed however that, contrary to our approach,
in such framework the speed of light is momentum dependent, so that not all phenomenological predictions
of those papers can be extended to our models.
Finally, it has been argued that, due to the modified dynamics, correction to the galaxy dynamics may
arise [20]. Also these, however, strongly depend on the specific realization of DSR in de Sitter space
9 Conclusions
It is known that DSR models can be derived from a 5-dimensional momentum space of coordinates πA ,
subject to the constraint πA
= −κ2 [17]. This is similar to the de Sitter constraint for the spacetime
coordinates. However, the physical interpretation is quite different. First of all, de Sitter spacetime inherits
a metric structure from the 5-dimensional space and this allows one to define a curvature. Different systems
of coordinates are physically equivalent. The momentum space, instead, does not possess a metric structure
and different coordinates cannot be considered physically equivalent, unless one adds further structure. In
fact, different realizations of DSR lead to different physical theories. Moreover, the mere existence of a de
Sitter group of transformations on a four-dimensional manifold does not automatically imply that this can
be identified with de Sitter space.
With these remarks in mind, one may try to construct a realization of a deformed de Sitter relativity
starting from five-dimensional space, similarly to what has been done for flat space [29]. Unfortunately,
however, it is not possible to impose contemporary constraints on the five-dimensional positions and momenta, and one is forced to start from a six-dimensional space. The construction of a Hamiltonian formalism in six-dimensional phase space with coordinates ΞM and momenta ΠM , subject to the constraints
Ξ2M = −α2 , Π2M = −κ2 has recently been investigated in [30].
From the study of DSR in de Sitter space one can also learn some lessons concerning the flat space
limit. First of all, it is useful to distinguish the translation generators, that dictate the conservation laws
for the momentum, from the physical momentum, identified with the phase space momentum variables.
This observation also gives a physical meaning to the auxiliary variables obeying canonical transformation
laws introduced in [15], whose interpretation was unclear: they are simply the generators of translations.
Moreover, it appears that the distinguishing feature of DSR is not the deformation of the Lorentz symmetry,
as usually postulated, but rather that of the translation symmetry, as exemplified by the Snyder model
discussed in Sects. 6 and 7 [9]. Of course, a complete discussion of this topic requires an operational
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
S. Mignemi: Doubly special relativity in de Sitter spacetime
definition of the momentum of a particle. In particular, for our interpretation to be valid, this should allow to
identify the measured value of the momentum with the canonical momentum rather than with the generator
of translations.
It would also be interesting to investigate the implication for cosmology of the energy dependence of
the cosmological constant, that could give rise to interesting deviations from standard picture at the early
stages of the evolution of the universe. This study may help to establish which one, among the infinite
possible realizations of DSR in de Sitter space some of which have been discussed in this paper, is closer
to physical reality.
Although we have not considered this subject in detail, it is also worth noticing that all our considerations can be easily extended to the case of anti-de Sitter space, by simply changing the sign of the
cosmological constant Λ.
I wish to thank J.P. Pereira for drawing my attention to [4].
M. Tegmark et al., Phys. Rev. D 69, 103501 (2004).
R. L. Mallett and G. N. Fleming, J. Math. Phys. 14, 45 (1973).
H. Y. Guo, C. G. Huang, Z. Xu, and B. Zhou, Phys. Lett. A 331, 1 (2004).
R. Aldrovandi, J. P. Beltrán Almeida, and J. P. Pereira, Class. Quantum Gravity 24, 1385 (2007).
S. Cacciatori, V. Gorini, and A. Kamenshchik, Ann. Phys. (Berlin) 17, 728 (2008).
G. Amelino-Camelia, Phys. Lett. B 510, 255 (2001); Int. J. Mod. Phys. D 11, 35 (2002).
G. Amelino-Camelia and T. Piran, Phys. Lett. B 497, 265 (2001); Phys. Rev. D 64, 036005 (2001).
H. S. Snyder, Phys. Rev. 71, 38 (1947);
J. Kowalski-Glikman and S. Nowak, Int. J. Mod. Phys. D 13, 299 (2003).
S. Mignemi, Phys. Lett. B 672, 186 (2009).
J. Lukierski, H. Ruegg, and W. J. Zakrzewski, Ann. Phys. (Berlin) 243, 90 (1995).
J. Magueijo and L. Smolin, Phys. Rev. Lett. 88, 190403 (2002); Phys. Rev. D 67, 044017 (2003).
J. Kowalski-Glikman, Mod. Phys. Lett. A 17, 1 (2002).
A. Granik, hep-th/0207113.
S. Mignemi, Phys. Rev. D 68, 065029 (2005).
S. Judes and M. Visser, Phys. Rev. D 68, 045001 (2003).
S. Ghosh and P. Pal, Phys. Rev. D 75, 105021 (2007).
J. Kowalski-Glikman, Phys. Lett. B 547, 291 (2002);
J. Kowalski-Glikman and S. Nowak, Class. Quantum Gravity 20, 4799 (2003).
S. Mignemi, Mod. Phys. Lett. A 18, 643 (2003).
A. Ballesteros, F. J. Herranz, and N. R. Bruno, hep-th/0409295.
J. Kowalski-Glikman and L. Smolin, Phys. Rev. D 70, 065020 (2004).
A. Marcianò et al., JCAP 1006, 030 (2010).
G. Amelino-Camelia, A. Marcianò, M. Matassa, and G. Rosati, arXiv:1006.0007.
S. Mignemi, Int. J. Mod. Phys. D 15, 925 (2006).
J. Magueijo and L. Smolin, Class. Quantum Gravity 21, 1725 (2004).
A. Kempf, G. Mangano, and R. B. Mann, Phys. Rev. D 52, 1108 (1995).
G. Amelino-Camelia, J. Kowalski-Glikman, G. Mandanici, and A. Procaccini, Int. J. Mod. Phys. A 20, 6007
D. Heyman, S. Major, and F. Hinteleitner, Phys. Rev. D 69, 105016 (2004).
S. Mignemi, Phys. Lett. A 373, 4401 (2009).
F. Girelli, T. Konopka, J. Kowalski-Glikman, and E. R. Livine, Phys. Rev. D 73, 045009 (2006).
S. Mignemi, Class. Quantum Gravity 26, 245020 (2009).
c 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Без категории
Размер файла
220 Кб
spacetime, relativity, sitter, double, special
Пожаловаться на содержимое документа