Kostadin Trencevski - Noncommutative Coordinates and Applications

Representation of the Lorentz Boosts in
6-dimensional Space-time
Kostadin Trenˇcevski
Institute of Mathematics, Ss. Cyril and Methodius University,
P.O.Box 162, Skopje, Macedonia, e-mail: kostatre@pmf.ukim.mk
1 Introduction
When we consider parallel transport of a 4-vector of veloc-
ity, the displaced 4-vector is again a 4-vector of velocity. But
if we consider the 4-vector of velocity as a Lorentz boost,
then its parallel displacement may not be a boost, because
may contain a space rotation, and can simultaneously give
information for both the velocity and space rotation of the
considered body. This is the main motivation for the present
paper, where we research a model of 3-dimensional time. The
evidence of the 3-dimensional time appears also in the quan-
tum mechanics, where besides the three spatial coordinate
operators appear 3 impulse coordinates, which are indeed
temporal coordinates.
Albert Einstein and Henri Poincare many years ago thought
about 3-dimensional time, such that the space and time would
be of the same dimension. At present time some of the au-
thors [1-5,7-9] propose multidimensional time in order to give
better explanation of the quantum mechanics and the spin.
1

2 Basic results
At each moment the set of all moving frames can be con-
sidered as a principal bundle over R3
, where the structural
group is the Lorentz group of transformations. This bundle
will be called space-time bundle. This bundle can be param-
eterized by the following 9 coordinates
{x, y, z}, {xs, ys, zs}, {xt, yt, zt},
such that the ﬁrst 6 coordinates parameterize the subbundle
with the ﬁber SO(3, R) . So this approach in the SR will be
called 3+3+3-dimensional model. Indeed, to each body are
related 3 coordinates for the position, 3 coordinates for the
space rotation and 3 coordinates to its velocity.
Firstly, we consider the analog of the Lorentz boosts from
the 3+1-dimensional space-time. The coordinates xs, ys, zs
and xt, yt, zt are functions of basic space coordinate x, y, and
z, and assume that the Jacobi matrices
V =
⎡
⎢
⎢
⎢
⎢
⎣
∂xs
∂x
∂xs
∂y
∂xs
∂z
∂ys
∂x
∂ys
∂y
∂ys
∂z
∂zs
∂x
∂zs
∂y
∂zs
∂z
⎤
⎥
⎥
⎥
⎥
⎦
and V ∗
=
⎡
⎢
⎢
⎢
⎢
⎣
∂xt
∂x
∂xt
∂y
∂xt
∂z
∂yt
∂x
∂yt
∂y
∂yt
∂z
∂zt
∂x
∂zt
∂y
∂zt
∂z
⎤
⎥
⎥
⎥
⎥
⎦
(2.1)
are respectively symmetric and antisymmetric. Further, let
us denote X = xs + ixt, Y = ys + iyt, Z = zs + izt, such
that the Jacobi matrix V = [∂(X,Y,Z)
∂(x,y,z) ] is Hermitian and V =
V + iV ∗
.
The antisymmetric matrix V ∗
depends on 3 variables and
2

its general form can be written as
V ∗
=
−1
c 1 − v2
c2
⎡
⎣
0 vz −vy
−vz 0 vx
vy −vx 0
⎤
⎦ . (2.2)
From (2.2) we can join to V ∗
a 3-vector v = (vx, vy, vz),
which transforms as a 3-vector. Namely, let we choose an
orthogonal 3×3 matrix P, which determines a space rotation
on the base B = R3
, applying to the coordinates x, y, z.
Then this transformation should also be applied to both sets
of coordinates {xs, ys, zs} and {xt, yt, zt}. Hence the matrix
V ∗
maps into PV ∗
P−1
= PV ∗
PT
, which corresponds to the
3-vector P · v. Thus v → P · v, and v is a 3-vector.
It is natural to assume that V should be presented in the
form
V = eiA
= cos A + i sin A.
Assume that A is an antisymmetric real matrix, which is
given by
A =
⎡
⎣
0 −k cos γ k cos β
k cos γ 0 −k cos α
−k cos β k cos α 0
⎤
⎦ ,
where v = c(cos α, cos β, cos γ) tanh(k) and (cos α, cos β, cos γ)
is a unit vector of the velocity vector. As a consequence we
obtain
sin A =
−1
c 1 − v2
c2
⎡
⎣
0 vz −vy
−vz 0 vx
vy −vx 0
⎤
⎦ , (2.3)
i.e. that V ∗
= sin A is given by (2.2), while the symmetric
3 × 3 matrix cos A is given by
(cos A)ij = V4δij +
1
1 + V4
ViVj, (2.4)
3

where (V1, V2, V3, V4) = 1
ic 1−v2
c2
(vx, vy, vz, ic).
From (2.1) and (2.2) the time vector in this special case is
given by
(xt, yt, zt) =
v
c 1 − v2
c2
× (x, y, z) + (x0
t , y0
t , z0
t ), (2.5)
where (x0
t , y0
t , z0
t ) does not depend on the basic coordinates.
The coordinates xt, yt, zt are independent and they cover the
Euclidean space R3
or an open subset of it. But the Ja-
cobi matrix [∂(xt,yt,zt)
∂(x,y,z)
] is a singular matrix as antisymmet-
ric matrix of order 3, where the 3-vector of velocity maps
into zero vector. So the quantity (xt, yt, zt) · v does not de-
pend on the basic coordinates and hence we assume that it
determines the 1-dimensional time t measured from
the basic coordinates. For example, if velocity is parallel
to the z-axis, then zt does not depend on the basic coordi-
nates because ∂zt
∂x
= ∂zt
∂y
= ∂zt
∂z
= 0 and hence zt is propor-
tional with the time from the basic coordinate system. Fur-
ther, one can easily verify that (1 − v2
c2 )−1/2
(v × (x, y, z)) =
(1 − v2
c2 )−1/2
(v × (cos A)−1
(xs, ys, zs)) = v × (xs, ys, zs) for si-
multaneous points in basic coordinates. So (2.5) becomes
(xt, yt, zt) =
v
c
× (xs, ys, zs) + c · Δt, (2.6)
where c = v
v · c.
4

3 Local isomorphism between O↑
+(1, 3) and SO(3, C)
Let us consider the following mapping F : O↑
+(1, 3) → SO(3, C)
given by
M 0
0 1
·
⎡
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎢
⎣
1 − 1
1+V4
V 2
1 − 1
1+V4
V1V2 − 1
1+V4
V1V3 V1
− 1
1+V4
V2V1 1 − 1
1+V4
V 2
2 − 1
1+V4
V2V3 V2
− 1
1+V4
V3V1 − 1
1+V4
V3V2 1 − 1
1+V4
V 2
3 V3
−V1 −V2 −V3 V4
⎤
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎥
⎦
→ M · (cos A + i sin A), (3.1)
where cos A and sin A are given by (2.4) and (2.3). This is
well defined because the decomposition of any matrix from
O↑
+(1, 3) as product of space rotation and a boost is unique.
Moreover, it is a bijection. In the following theorem is con-
structed effectively such an isomorphism [10].
Theorem 1. The mapping (3.1) defines (local) isomor-
phism between the groups O↑
+(1, 3) and SO(3, C).
Indeed, the mapping
⎡
⎢
⎢
⎣
0 c −b ix
−c 0 a iy
b −a 0 iz
−ix −iy −iz 0
⎤
⎥
⎥
⎦ →
⎡
⎣
0 c + iz −b − iy
−c − iz 0 a + ix
b + iy −a − ix 0
⎤
⎦
defines an isomorphism between the Lie algebras o(1, 3) and
o(3, C). This isomorphism induces local isomorphism be-
tween O↑
+(1, 3) and SO(3, C), and it induces (local) isomor-
phism between the two groups. Further it is proved that this
(local) isomorphism is given by (3.1).
If we want to find the composition of two space-time trans-
formations which determine space rotations and velocities,
5

there are two possibilities which lead to the same result: to
multiply the corresponding two matrices from SO(3, C) or
from O↑
+(1, 3). Since the result is the same, the three dimen-
sionality of the time is difficult to detect, and we feel like the
time is 1-dimensional. The essential difference in using these
two methods is the following. The Lorentz transformations
give relationship between the coordinates of a 4-vector with
respect tSo they show how the coordinates of a considered 4-
vector change by changing the base space. On the other side,
the matrices of the isomorphic group SO(3, C) show how the
space rotation and velocity change between two bodies, us-
ing the chosen base space, by consideration of changes in the
fiber. So we have a duality in the SR. The use of the group
SO(3, C) alone is not sufficient, because their matrices are
only Jacobi matrices free from any motion.
4 Preparation for the Theorem 3
(i) In the next section we want to deduce the Lorentz trans-
formations using the group SO(3, C). We assume that there
is no effective motion, i.e. change of the basic coordinates,
but simply rotation for an imaginary angle. Such a trans-
formation will be called passive motion. The examination
of observation of a moving body can easily be done in the
following way.
Let us assume that vx = v, vy = vz = 0. In this case the
matrix V = cos A determined by (2.4) is given by
V = cos A = diag(1,
1
1 − v2
c2
,
1
1 − v2
c2
).
Hence there is no length contraction in the direction of mo-
tion (x-direction), while the lengths in any direction orthog-
6

onal to the direction of motion (yz-plane) are observed to be
larger (1 − v2
c2 )
−1/2
times. Notice that if we multiply all these
length coefficients by 1 − v2
c2 we obtain the prediction from
the SR. Hence the observations for lengths for passive and
active motions together is in agreement with the classical
known results.
If there is an active motion, i.e. there is change of the basic
coordinates, independently from the previous effect, we have
the following phenomena. The group SO(3, C) preserves the
quantity (Δx + icΔtx)2
+ (Δy + icΔty)2
+ (Δz + icΔtz)2
and
hence also preserves Δx2
+Δy2
+Δz2
−c2
(Δt2
x+Δt2
y +Δt2
z) =
Δx2
+ Δy2
+ Δz2
− c2
Δt2
.
(ii) The previous conclusion for the spatial lengths can be
supported by the following conclusion about time intervals.
While the observation of lengths may be done in different
directions, the observation of time flow does not depend on
the direction, but only on velocity. The time observed in
a moving system is slower for coefficient 1 − v2
c2 for active
motions (analogously to the lengths). It is a consequence
of the relativistic law of adding collinear velocities and it is
presented by the following theorem ([10]).
Theorem 2. Assume that the relativistic law of summa-
tion of collinear velocities is satisfied, and assume that the
observed time in a moving inertial coordinate system with
velocity v is observed to be multiplied with f(v
c), where f is a
differentiable function and the first order Taylor development
of f does not contain linear summand of v/c. Then, f must
be f(v
c ) = 1 − v2
c2 .
Since the 1-dimensional time direction is parallel to the
velocity vector, there is no change in the observation of the
7

time vector which corresponds to the passive motion. So
the observed change for the time vector considered in the
previous theorem comes only from the active motion.
Using the Theorem 2 and the assumption that the 1 - di-
mensional time is a quotient between the 3-vector of displace-
ment and the 3-vector of velocity, the following conclusion is
deduced in [10]. Let the initial and the end point of a 4-
vector r be simultaneous in one coordinate system S . Then
these two points in another coordinate system diﬀer for time
δt =
r v
c2
1 − v2
c2
, (4.1)
where v is the velocity vector. Notice that (4.1) is also a
consequence from the Lorentz transformations.
(iii) The base manifold R3
is 3-dimensional. It is conve-
nient to consider it as a subset of C3
, consisting of
(x, y, z, ctx, cty, ctz),
where ctx = cty = ctz = 0 at a chosen initial moment, and
call it complex base. The change of the coordinates can be
done via the 6×6 real matrix
M cos A −M sin A
M sin A M cos A
, where
M is a space rotation. It acts on the 6-dimensional vectors
(Δx, Δy, Δz, 0, 0, 0)T
of the introduced complex base. Multi-
plying the vectors of the complex base (Δx, Δy, Δz, 0, 0, 0)T
from left with this matrix, we obtain 3-dimensional base
subspaces as they are viewed from the observer who rests
with respect to the chosen complex base. Moreover, the
pair ((Δx, Δy, Δz, 0, 0, 0)T
, G) ∈ R6
× SO(3, C) viewed for
moving and rotated base space determined by the matrix
P ∈ SO(3, C) is given by
(P(Δx, Δy, Δz, 0, 0, 0)T
, PGPT
) ∈ R6
× SO(3, C).
8

(iv) Until now we considered mainly the passive motions,
while our goal is to consider active motion in the basic coor-
dinates. The active motion is simply translation in the basic
space, caused by the ﬂow of the time. So besides the complex
rotations of SO(3, C) we should consider also translations in
C3
. Now (Δctx, Δcty, Δctz) for the basic coordinates is not
more a zero vector. The time which can be measured in basic
coordinates is Δt = [(Δtx)2
+ (Δty)2
+ (Δtz)2
]
1/2
. In case of
motion of a point with velocity v we have translation in the
basic coordinates for the vector vΔt + icΔt. The space part
vΔt is obvious, while the time part cΔt follows from (2.6).
An orthogonal complex transformation may be applied, if
previously the basic coordinates are translated.
5 Lorentz transformations as transformations on C3
For the sake of simplicity we will omit the symbol ”Δ” for
space coordinates. Let the coordinates xs, ys, zs are denoted
by x , y , z and let us denote r = (x, y, z) and r = (x , y , z ).
It is of interest to see the form of the Lorentz boosts as
transformations in C3
, while the space rotations are identical
in both cases.
Theorem 3. The following transformation in C3
(1−
v2
c2
)
−1/2 r
ct + v×r
c
=
cos A − sin A
sin A cos A
r + v(t + δt)
c(t + δt)
(5.1)
via the group SO(3, C) is equivalent to the transformation
given by a Lorentz boost determined by the isomorphism (3.1).
The coeﬃcient β = (1 − v2
c2 )−1/2
is caused by the active
motion (i). There is a translation in the basic coordinates
9

for vector (v(t + δt), c(t + δt)), where δt is defined by (4.1).
On the other side, according to (2.6) in the moving system
we have the time vector v × r /c, which disappears in basic
coordinates (v = 0).
Proof. Notice that if we consider a space rotation P, which
applies to all triples, the system (5.1) remains covariant. In-
deed, r, r , v, c, v × r transform as vectors, t and δt, which is
defined by (4.1), transform as scalars, while cos A and sin A
transform as tensors of rank 2. Hence, if we multiply from
left with
P O
O P
the both sides of (5.1), we obtain
β
Pr
Pct + (P v)×(P r )
c
=
=
P cos APT
−P sin APT
P sin APT
P cos APT
Pr + Pv(t + δt)
Pc(t + δt)
and since
P(cos A)PT
= cos(PAPT
) and P(sin A)PT
= sin(PAPT
),
the covariance of (5.1) is proved. So it is sufficient to apply
such a transformation P which maps vector v into (v, 0, 0)
and to prove the theorem in this special case.
Notice that both left and right side of (5.1) are linear func-
tions of x, y, z, t, x , y , z , t , and so after some transforma-
tions it can be simplified. Then the first three equations of
(5.1) reduce to the following three equations respectively
x =
x + vt
1 − v2
c2
, y = y, z = z.
Further, using these three equations, the fourth equation of
(5.1) reduces to
t =
t + vx
c2
1 − v2
c2
,
10

while the 5-th and the 6-th equations are identically satisfied.
2
According to Theorem 3 the well known 4-dimensional
space-time is not fixed in 6 dimensions, but changes with
the direction of velocity. Namely this 4-dimensional space-
time is generated by the basic space vectors and the velocity
vector from the imaginary part of the complex base.
Having in mind the equation (2.6), the Lorentz transfor-
mation (5.1) can be written in the following form
(1 −
v2
c2
)
−1/2 rs
rt
=
cos A − sin A
sin A cos A
r + v(t + δt)
c(t + δt)
.
(5.2)
The coefficient β = (1−v2
c2 )−1/2
is caused by the active motion
(i). Since the coordinates xs, ys, zs, xt, yt, zt are measured ac-
cording to the basic coordinates x, y, z, we know that all of
them are observed contracted for coefficient 1 − v2
c2 so we
needed to multiply them with β = (1 − v2
c2 )−1/2
. If we denote
again the same coordinates but now measured from the self
coordinate system, then the Lorentz transformation becomes
rs
rt
=
cos A − sin A
sin A cos A
r + v(t + δt)
c(t + δt)
. (5.3)
Analogous result can be obtained in case if there is simulta-
neously motion with a velocity and space rotation. If there
is also rotation determined by the matrix P ∈ SO(3, R),
then the matrix
cos A − sin A
sin A cos A
should be replaced by
P cos A −P sin A
P sin A P cos A
.
11

6 Cosmology based on the 3+3+3-model
According to (5.3) both vectors
rs
rt
and
r + v(t + δt)
c(t + δt)
lie
in the same 6-dimensional space. This is true, if the space-
time of the Universe globally is Euclidean space with dimen-
sion 6. But, however, this is not in accordance with the
temporary cosmology. Neglecting the time coordinates, it is
accepted to be a 3-sphere. We shall modify this statement,
by identifying the antipode points of the 3-sphere, and hence
obtaining the projective space RP3
. This space is homeo-
morphic with the Lie group SO(3, R). The local coordinates
of SO(3, R) are angles, i.e. real numbers, but we use length
units for our local space coordinates. So for each small an-
gle ϕ of rotation in a given direction corresponds coordinate
length Rϕ in the same direction, where R is a constant which
can be called radius of the Universe. By accepting this mod-
ification of the spatial part of the Universe, we do not change
anything locally, because the 3-sphere has locally the group
structure of the unit quaternions, and locally this group is
isomorphic with the group SO(3, R). According to this small
modification of the Universe it is now natural to assume that
the space time of the Universe is isomorphic to SO(3, C).
Notice that analogously as (x+ictx, y+icty, z+ictz) is a lo-
cal coordinate neighborhood of SO(3, C), i.e. the space-time
of the Universe, also (1 − v2
c2 )
−1/2
(xs + ixt, ys + iyt, xz + izt) is
a coordinate neighborhood of the same manifold. These two
coordinate systems can be considered as coordinate neigh-
borhoods of SO(3, C) as a complex manifold, because ac-
cording to (5.3) the Cauchy-Riemannian conditions for these
two systems are satisfied.
Let us discuss the space-time dimensionality of the Uni-
12

verse. We mentioned at the beginning of this section that
the dimensionality is 6 if the space-time is ﬂat. But now
we have that it is parameterized by the following 9 inde-
pendent coordinates: x, y, z coordinates which locally pa-
rameterize the spatial part of the Universe SO(3, R), and
xs, ys, zs, xt, yt, zt coordinates which parameterize the bun-
dle. Dually, the partial derivatives of these coordinates with
respect to x, y, z lead to the same manifold, but now as a
group of transformations. So in any case the total space-time
of the Universe is homeomorphic to SO(3, R)×SO(3, C), i.e.
SO(3, R) × R3
× SO(3, R). In the above parameterization,
R3
is indeed the space of velocities such that |v| < c. More-
over, the group SO(3, C) as well as the isomorphic group
O↑
+(1, 3) considers only velocities with magnitude less than
c. If |v| = c, then we have a singularity.
Notice that if we know the coordinates x, y, z and also
xs, ys, zs, xt, yt, zt, then according to the Lorentz transforma-
tions, the time coordinates cxt, cyt, czt are uniquely deter-
mine. And conversely, if we know the coordinates xs, ys, zs,
and also x, y, z, ctx, cty, ctz, then the time coordinates xt, yt, zt
are uniquely determined. So we can say that there are 6
spatial and 3 temporal coordinates. Notice that if we
consider that the Universe is a set of points, then it is more
natural to consider it as 6-dimensional. But, since we con-
sider the Universe as a set of orthonormal frames, so it is
more natural to consider it as 9-dimensional.
This 9-dimensional space-time has the following property:
From each point of the space-time, each velocity and
each spatial direction of the observer, the Universe
seems to be the same. In other words, there is no priv-
ileged space points (assuming that R is a global constant),
no privileged direction and no privileged velocity. In other
13

words, everything is relative.
Finally we can conclude the following. In the (total) space-
time SO(3, R) × SO(3, C), i.e. SO(3, R) × R3
× SO(3, R)
there are three essential fiber bundles: (i) First case where
the base is the space part of the Universe SO(3, R) and the
fiber is the group of orthogonal transformations SO(3, C),
which consists of all rotations and motions with velocities;
(ii) second case where the base is the set of all spatial rota-
tions SO(3, R), and the fiber is the space-time part of the
Universe SO(3, C); (iii) third case where the base is the time
part of the Universe R3
and the fiber is the set of space part of
the universe and all space rotations, i.e. SO(3, R)×SO(3, R).
The first and the second fiber bundles are direct products
of the base SO(3, R) and the fiber SO(3, C). The first bun-
dle was studied in many details in the first sections, while
the second bundle was studied mainly in this section, after
the topological structure of the space-time of the Universe
was accepted. Here we want to emphasize that if we want to
consider the fiber over a point B instead of the fiber over a
point A (case (i)), then there is unique matrix P ∈ SO(3, R),
which maps the point A into point B, because both points
A and B are matrices. In this case each matrix M from the
fiber SO(3, C) should be replaced with PMPT
. This implies
that each 3-vector of velocity v should be replaced by Pv and
each space rotation M should be replaced by PMPT
. If A
and B are relatively close points in the Universe, then P is
practically unit matrix, and each matrix in the fiber remains
unchanged. This is usually interpreted as translations in flat
space-time as in Minkowski space. This case is close to the
methods of the classical and relativistic mechanics. Indeed,
it is sufficient to study the law of the change of the matrix
M from the fiber, i.e. the matrix which consists the infor-
14

mations about the spatial rotation and the velocity vector of
a considered test body. Then it is easy to find the trajectory
of motion of the test body.
Now let us consider the case (ii). Let us choose an ar-
bitrary space rotation. This may be done by arbitrary 3
orthonormal tangent vectors over the Universe considered as
SO(3, R) manifold, and then it can be transfered at each
point of SO(3, R), by using the group structure of SO(3, R).
Then the studying of motion of arbitrary test body means
to find how changes the matrix which gives the position of
the test body in the space and its velocity at the chosen
moment. The space rotation of the test body also changes,
analogously as the position of the test body in case (i) also
changes, but in case (ii) we are not interesting about it. This
is close to the methods of the quantum mechanics, where the
coordinate operators (i.e. space coordinate operators) and
the impulse operators (i.e. the time coordinate operators)
have the main role. If we want to use another space rotation
instead of the chosen one, then there exists a unique matrix
P which maps the initial space rotation into the new space
rotation. In this case each matrix M from the fiber SO(3, C)
should be replaced with PMPT
. This implies that each 3-
vector of velocity v should be replaced by Pv and each space
position, which is given by orthogonal matrix S, should be
replaced by PSPT
. Since the radius of the Universe is ex-
tremely large, practically for each space position, the matrix
S for space position is close to the unit matrix, and approx-
imately can be considered as a vector of translation r, and
now this vector should be replaced by Pr.
15

While in the first and the second case the group and the
fiber is simultaneously SO(3, C), in the third case the fiber
and the group of the fiber is SO(3, R) × SO(3, R), which is
given by (X, Y )(X , Y ) = (XX , Y Y ). So, while the time
coordinates are functionally dependnt with the both sets of
spatial coordinates, both sets of spatial coordinates mutually
independent.
The previous discussion can be summarized by the follow-
ing diagram,
V ∼= R3
× ×
S ∼= SO(3, R) × SR ∼= SO(3, R)
consisting of three 3-dimensional sets: velocity (V) which is
homeomorphic to R3
, space (S) which is homeomorphic to
SO(3, R), and space rotation (SR) which is homeomorphic
to SO(3, R). Each of these three sets can be considered as
base, while the Cartesian product of the other two sets can be
considered simultaneously as a fiber and also a group which
acts over the fiber. The first set is not a group and must be
joined with each of the other two sets, while the other two
sets may exist independently because they are groups.
7 Modification of the Lorentz transformations
By acceptance a priori of the previous assumptions, as a
consequence we do not have any translations and vectors.
So the classical Lorentz transformations have no sense any
more. From this viewpoint, we should modify the Lorentz
transformation defined on a flat space. The vectors
rs
rt
16

and
r + v(t + δt)
c(t + δt)
should be replaced by two matrices and
the Lorentz transformation will become a matrix equality in
the following way. Notice that each small neighborhood of
the unit matrix in SO(3, C) can be parameterized by the
following 6-dimensional vector (x, y, z, x , y , z ), i.e. vector
(z1, z2, z3) ∈ C3
, where z1 = x+ ix , z2 = y + iy , z3 = z + iz ,
by joining the following matrix in SO(3, C)
⎧
⎨
⎩
1
Δ
⎡
⎣
1 + z2
1 − z2
2 − z2
3 −2z3 + 2z1z2 2z2 + 2z1z3
2z3 + 2z1z2 1 − z2
1 + z2
2 − z2
3 −2z1 + 2z2z3
−2z2 + 2z1z3 2z1 + 2z2z3 1 − z2
1 − z2
2 + z2
3
⎤
⎦
⎫
⎬
⎭
1/2
,
where Δ = 1 + z2
1 + z2
2 + z2
3. In a special case, if x =
y = z = 0, this matrix represents a spatial rotation in the
direction of (x, y, z) = (z1, z2, z3) and the angle of rotation
is ϕ = arctan x2 + y2 + z2. This representation can be
extended to the following special case, which will be used
later. If x
x
= y
y
= z
z
= v
c
, then the direction of rotation
is (x, y, z) = v
c(x , y , z ) and the complex angle of rotation
satisﬁes
tan ϕ = (x + ix )2 + (y + iy )2 + (z + iz )2 =
= x2 + y2 + z2(1 + i
c
v
).
For example, if v = (0, 0, v), then the corresponding orthog-
onal matrix is given by
⎡
⎣
cos z
R
− sin z
R
0
sin z
R cos z
R 0
0 0 1
⎤
⎦
⎡
⎣
cosh z
R −i sinh z
R 0
i sinh z
R
cosh z
R
0
0 0 1
⎤
⎦ =
=
⎡
⎣
cosh z
R
−i sinh z
R
0
i sinh z
R cosh z
R 0
0 0 1
⎤
⎦
⎡
⎣
cos z
R − sin z
R 0
sin z
R
cos z
R
0
0 0 1
⎤
⎦ . (7.1)
17

The equality (5.2) now can be represented in the following
way. First note that (5.2) can be written in the following
form
1
R
rs
rt
=
1
R
cos A − sin A
sin A cos A
r
0
1 −
v2
c2
+
+
1
R
v
c
(t + δt) 1 −
v2
c2
.
Both vectors 1
R
rs
rt
and 1
R
cos A − sin A
sin A cos A
r
0
1 − v2
c2
should be replaced by two matrices M and M from SO(3, C),
while the vector 1
R
v
c
(t + δt) 1 − v2
c2 should be replaced
by the following 3 × 3 complex orthogonal matrix
L
⎡
⎢
⎢
⎣
cos
v(t+δt) 1−v2
c2
R
− sin
v(t+δt) 1−v2
c2
R
0
sin
v(t+δt) 1−v2
c2
R
cos
v(t+δt) 1−v2
c2
R
0
0 0 1
⎤
⎥
⎥
⎦ ×
×
⎡
⎢
⎢
⎣
cosh
c(t+δt) 1−v2
c2
R
−i sinh
c(t+δt) 1−v2
c2
R
0
i sinh
c(t+δt) 1−v2
c2
R
cosh
c(t+δt) 1−v2
c2
R
0
0 0 1
⎤
⎥
⎥
⎦ LT
,
which obtains from (7.1) by replacing z = v(t + δt) 1 − v2
c2
and z = c(t + δt) 1 − v2
c2 , and where L ∈ SO(3, R) is ar-
bitrary orthogonal matrix which maps the vector (0, 0, 1)T
into the vector (vx/v, vy/v, vz/v)T
. Hence the Lorentz trans-
18

formation (5.2) takes the following matrix from
M = ML
⎡
⎢
⎢
⎣
cos
v(t+δt) 1−v2
c2
R − sin
v(t+δt) 1−v2
c2
R 0
sin
v(t+δt) 1−v2
c2
R
cos
v(t+δt) 1−v2
c2
R
0
0 0 1
⎤
⎥
⎥
⎦ ×
×
⎡
⎢
⎢
⎣
cosh
c(t+δt) 1−v2
c2
R −i sinh
c(t+δt) 1−v2
c2
R 0
i sinh
c(t+δt) 1−v2
c2
R
cosh
c(t+δt) 1−v2
c2
R
0
0 0 1
⎤
⎥
⎥
⎦ LT
, (7.2)
and now there are no translations. Notice that (7.2) is not
equivalent to (5.2). But neglecting the terms of order R−2
and less, which are extremely small, the equalities (5.2) and
(7.2) are equivalent. However, we accept now that (7.2) is
exact equality, while (5.2) is approximative.
References
[1] V. S. Barashenkov, Multitime generalization of Maxwell
electrodynamics gravity, Tr. J. of Phys., 23, 1999, 831-
838.
[2] V. S. Barashenkov, Quantum ﬁeld theory with three-
dimensional vector time, Particles and Nuclei, Letters,
2, 2004, 119.
[3] V. S. Barashenkov and M. Z. Yuriev, Solutions of mul-
titme Dirac equations, Particles and Nuclei, Letters, 6,
2002, 115.
[4] E. A. B. Cole, Particle decay in six-dimensional relativ-
ity, J. Phys. A: Math. Gen., 1980, 109-115.
19

[5] E. A. B. Cole, Comments on the use of three time di-
mensions in Relativity, Phys. Lett., 76A, 1980, 371.
[6] A. Franco R. Jorge, Vectorial Lorentz transformations,
Elec. J. Theor. Phys., 9, 2006, 35-64.
[7] H. Kitada, Theory of local times, Nuovo Cim. B, 109,
1994, 281-202.
[8] J. Strnad, Once more on multi-dimensional time, J.
Phys. A: Math. Gen., 14, 1981, L433-L435.
[9] J. Strnad, Experimental evidence against three-
dimensional time, Phys. Lett., 96A, 1983, 371.
[10] K. Trenˇcevski, Duality in the Special Relativity based on
the isomorphic groups SO(3, C) and O↑
+(1, 3), submitted
for publication in Tensor.
[11] K. Trenˇcevski, E. Celakoska, V. Balan, On a parameter-
ization of the Poincare group, Int. J. of pure and Appl.
Math., 30 No. 1 (2006), 109-117.
20

Kostadin Trencevski - Noncommutative Coordinates and Applications

More Related Content

What's hot (19)

Viewers also liked (9)

Similar to Kostadin Trencevski - Noncommutative Coordinates and Applications (20)

More from SEENET-MTP (20)

Recently uploaded (20)

Kostadin Trencevski - Noncommutative Coordinates and Applications