, ,

1 Introduction

The starting point of noncommutative geometry is the passage from geometric spaces to algebras of functions on these spaces with the subsequent translation of basic analytic and geometric concepts and constructions on geometric spaces to the algebraic language and their extension to general noncommutative algebras. Such a procedure is well-known and was applied for a long time, for instance, in algebraic geometry, where it is related with the study of commutative algebras. It is also well known that the theory of

C^{*}

-algebras is a far-reaching generalization of the theory of topological spaces, and the theory of von Neumann algebras is a generalization of the classical measure and integration theory.

The main purpose of noncommutative differential geometry, which was initiated by Connes [39] and is actively developing at present time (cf. the recent surveys [45, 46] and the books [42, 81, 119] in regard to different aspects of noncommutative geometry), consists in the extension of the methods described above to the analytic objects on geometric spaces and to the noncommutative algebras. Here the main attention is focused on that, first, a correct noncommutative generalization applied in the classical setting, that is, to an algebra of functions on a compact manifold should agree with its classical analogue, and, second, it should inherit nice algebraic and analytic properties of its classical analogue. Nevertheless, it should be noted that, as a rule, such noncommutative generalizations are quite nontrivial and have richer structure and essentially new features than their commutative analogues.

Noncommutative geometry lies on the border of functional analysis and differential geometry and is of great importance for these areas of mathematics.

On the one hand, the development of geometric methods in the operator theory and the theory of operator algebras allows to use fruitfully geometric intuition for the investigation of various problems of abstract functional analysis. On the other hand, there are many examples of singular geometric spaces, which are badly described from the point of view of classical measure theory and resist to the study by usual “commutative” methods of geometry, topology and analysis, but one can naturally associate to them a noncommutative algebra, which can be considered as an analogue of the algebra of (measurable, continuous, smooth and so on) functions on the given geometric object. Let us give some examples of such singular objects:

(1) Manifolds with singularities, for instance, manifolds with isolated conic singular points.
(2) Discrete spaces.
(3) The dual space to a group (discrete or a Lie group).
(4) Cantor sets.
(5) The orbit space of a group action on a manifold.
(6) The leaf space of a foliation.

Use of the notions and methods of noncommutative geometry for a noncommutative algebra, being an analogue of an algebra of functions on a singular geometric space, allows in many cases, first of all, just to define some reasonable analytic and geometric objects associated with the given space and pose sensible problems, that, in its turn, gives possibility to apply properties of these objects for getting an information about geometry of this space.

This paper contains a short exposition of the methods of noncommutative geometry as applied to the study of one of the fundamental examples of noncommutative geometry, namely, the leaf spaces of a foliation on a smooth manifold, or, that is the same, to the study of the transverse structure of foliations. Our purpose is to give a survey of the basic notions and methods of noncommutative geometry, to show how one can associate various objects of noncommutative geometry to foliations and how these noncommutative analogues of classical notions are related with classical objects on foliated manifolds, and to describe applications of the methods of noncommutative geometry to the study of geometry of foliations.

The author is grateful to N.I. Zhukova for useful remarks.

2 Background information on foliation theory

2.1 Foliations

In this Section, we recall the definition of a foliated manifold and some notions related to foliations (concerning to different aspects of the foliation theory see, for instance, [28, 29, 30, 74, 133, 136, 137, 153, 177] ).

Let

M

be a smooth manifold of dimension

n

. (Here and later on, “smooth” means “of class

C^{\infty}

”. We will always assume that all our objects under consideration are of class

C^{\infty}

Definition 2.1. (1) An atlas

A = {(U_{i}, φ_{i})}

, where

φ_{i} : U_{i} \subset M \to R^{n}

, of the manifold

M

is called an atlas of a foliation of dimension

p

and codimension

q

(

p \leq n, p + q = n

), if, for any

i

and

j

such that

U_{i} \cap U_{j} \neq \emptyset

, the coordinate transformation

φ_{i j} = φ_{i} \circ φ_{j}^{- 1} : φ_{j} (U_{i} \cap U_{j}) \subset R^{p} \times R^{q} \to φ_{i} (U_{i} \cap U_{j}) \subset R^{p} \times R^{q}

has the form

φ_{i j} (x, y) = (α_{i j} (x, y), γ_{i j} (y)), (x, y) \in φ_{j} (U_{i} \cap U_{j}) \subset R^{p} \times R^{q} .

(2) Two atlases of a dimension

p

foliation are equivalent, if their union is again an atlas of a dimension

p

foliation.

(3) A manifold

M

endowed with an equivalence class

ℱ

of atlases of a dimension

p

foliation is called a manifolds with a dimension

p

foliation (or a foliated manifold).

An equivalence class

ℱ

of foliation atlases is also called a complete atlas of a foliation. We will also say that

ℱ

is a foliation on

M

A pair

(U, φ)

that belongs to the atlas of the foliation

ℱ

and also the corresponding map

φ : U \to R^{n}

are called a foliated chart of the foliation

ℱ

, and

U

is called a foliated coordinate neighborhood.

Let

φ : U \subset M \to R^{n}

be a foliated chart. The connected components of the set

φ^{- 1} (R^{p} \times {y}), y \in R^{q}

are called the plaques of the foliation

ℱ

Plaques of

ℱ

given by all possible foliated charts form a base of a topology on

M

. This topology is called the leaf topology on

M

. We will also denote by

ℱ

the set

M

endowed with the leaf topology. One can introduce the structure of a

p

-dimensional manifold on

ℱ

Connected components of the manifold

ℱ

are called leaves of the foliation

ℱ

. Leaves are (one-to-one) immersed

p

-dimensional submanifolds in

M

. For any

x \in M

, there is a unique leaf, passing through

x

. We will denote this leaf by

L_{x}

One can give the equivalent definition of a foliation, saying, that there is given a foliation

ℱ

of dimension

p

on a manifold

M

of dimension

n

, if

M

is represented as a disjoint union of a family

{L_{λ} : λ \in ℒ}

of connected, (one-to-one) immersed submanifolds of dimension

p

, and there is an atlas

A = {(U_{i}, φ_{i})}

of the manifold

M

such that, for any

i

and for any

λ \in ℒ

, the connected components of the set

L_{λ} \cap U_{i}

are given by equations of the form

y = const

Example 2.2. Let

M

be a smooth manifold of dimension

n

B

a smooth manifold of dimension

q

and

π : M \to B

a submersion (that is, the differential

d π_{x} : T_{x} M \to T_{π (x)} B

is surjective for any

x \in M

). The connected components of the pre-images of points of

B

under the map

π

determine a codimension

q

foliation on

M

, which is called the foliation determined by the submersion

π

. If, in addition, the pre-images

π^{- 1} (b), b \in B,

are connected, the foliation is called simple.

Example 2.3. Let

X

be a nonsingular (that is, a non-vanishing) smooth vector field on a compact manifold

M

. Then its phase curves form a codimension one foliation.

More generally, suppose that a connected Lie group

G

acts smoothly on a smooth manifold

M

, and, moreover, the dimension of the isotropy subgroup

G_{x} = {x \in G : g x = x}

does not depend on

x \in M

. In particular, one can assume that the action is locally free, that means

G_{x}

is discrete for any

x \in M

. Then the orbits of the Lie group

G

action define a foliation on

M

Example 2.4. Linear foliation on torus. Consider a vector field

\tilde{X}

R^{2}

given by

\tilde{X} = α \frac{\partial}{\partial x} + β \frac{\partial}{\partial y}

with constant

α

and

β

. Since

\tilde{X}

is invariant under all translations, it determines a vector field

X

on the two-dimensional torus

T^{2} = R^{2} / Z^{2}

. The vector field

X

determines a foliation

ℱ

T^{2}

. The leaves of

ℱ

are the images of the parallel lines

\tilde{L} = {(x_{0} + t α, y_{0} + t β) : t \in R}

with the slope

θ = β / α

under the projection

R^{2} \to T^{2}

In the case when

θ

is rational, all leaves of

ℱ

are closed and are circles, and the foliation

ℱ

is determined by the fibers of a fibration

T^{2} \to S^{1}

. In the case when

θ

is irrational, all leaves of

ℱ

are everywhere dense in

T^{2}

Example 2.5. Homogeneous foliations. Let

G

be a Lie group and

H \subset G

its connected Lie subgroup. The family

{g H : g \in G}

of right cosets of

H

forms a foliation

ℋ

G

. If

H

is a closed subgroup, then

G / H

is a manifold and

ℋ

is a foliation, whose leaves are given by the fibers of the fibration

π : G \to G / H

Moreover, suppose that

Γ \subset G

is a discrete subgroup

G

. Then the set

M = Γ \ G

of left cosets of

Γ

is a manifold of the same dimension as

G

. If

Γ

is cocompact in

G

M

is compact. In any case, because

ℋ

is invariant under the left translations, and

Γ

acts on the left, the foliation

ℋ

is mapped by the map

G \to M = Γ \ G

to a well-defined foliation

ℋ_{Γ}

M

, which is often denoted by

ℱ (G, H, Γ)

and is called a locally homogeneous foliation. The leaf of

ℋ_{Γ}

through a point

Γ g \in M

is diffeomorphic to

H / (g Γ g^{- 1} \cap H)

Example 2.6. The Reeb foliation. Let us describe a classical construction of a codimension 1 foliation on the three-dimensional sphere

S^{3}

due to Reeb [152] . Let

D^{2}

denote the disk

{(x, y) \in R^{2} : x^{2} + y^{2} \leq 1}

. We start with the foliation in the cylinder

{(x, y, z) \in R^{3} : x^{2} + y^{2} \leq 1} = D^{2} \times R

, whose leaves are the boundary of the cylinder

x^{2} + y^{2} = 1

and surfaces

z = c + exp (1 / (1 - x^{2} - y^{2}))

with an arbitrary constant

c

. Since this foliation is invariant under translations in

z

, it is mapped by the standard projection

D^{2} \times R \to D^{2} \times R / Z = D^{2} \times S^{1}

to a foliation

ℱ_{R}

on the solid torus

D^{2} \times S^{1}

. Finally, observe that the standard three-dimensional sphere

S^{3} = {x = (x_{1}, x_{2}, x_{3}, x_{4}) \in R^{4} : x_{1}^{2} + x_{2}^{2} + x_{3}^{2} + x_{4}^{2} = 1}

is obtained by gluing along the boundary of two copies of

D^{2} \times S^{1}

. More precisely,

S^{3} = S_{1}^{3} \cup S_{2}^{3}

, where

S_{1}^{3} = {x \in S^{3} : x_{1}^{2} + x_{2}^{2} \leq x_{3}^{2} + x_{4}^{2}}, S_{2}^{3} = {x \in S^{3} : x_{1}^{2} + x_{2}^{2} \geq x_{3}^{2} + x_{4}^{2}} .

A diffeomorphism

S_{1}^{3} \overset{\sim =}{⟶} D^{2} \times S^{1}

is given by

x \in S_{1}^{3} \mapsto (\frac{x_{1}}{\sqrt{x_{3}^{2} + x_{4}^{2}}}, \frac{x_{2}}{\sqrt{x_{3}^{2} + x_{4}^{2}}}, \frac{x_{3}}{\sqrt{x_{3}^{2} + x_{4}^{2}}}, \frac{x_{4}}{\sqrt{x_{3}^{2} + x_{4}^{2}}}) \in D^{2} \times S^{1},

where we consider

S^{1}

{(x, y) \in R^{2} : x^{2} + y^{2} = 1}

, and its inverse has the form

(y_{1}, y_{2}, y_{3}, y_{4}) \in D^{2} \times S^{1} \mapsto (\frac{y_{1}}{| y |}, \frac{y_{2}}{| y |}, \frac{y_{3}}{| y |}, \frac{y_{4}}{| y |}) \in S_{1}^{3},

where

| y |^{2} = y_{1}^{2} + y_{2}^{2} + y_{3}^{2} + y_{4}^{2}

. Similar formulas can be written for

S_{2}^{3}

Take the Reeb foliations

ℱ_{R}

S_{1}^{3}

and

S_{2}^{3}

. Since the boundaries

\partial S_{1}^{3}

and

\partial S_{2}^{3}

are compact leaves, there is a well-defined foliation on the union of

S_{1}^{3}

and

S_{2}^{3}

, that is, on the sphere

S^{3}

, which is called the Reeb foliation on

S^{3}

. The only compact leaf of the Reeb foliation is the common boundary

\partial S_{1}^{3} = \partial S_{2}^{3}

S_{1}^{3}

and

S_{2}^{3}

, diffeomorphic to the two-dimensional torus

T^{2}

. The other leaves are diffeomorphic to the plane

R^{2}

Example 2.7. Suspension. Let

B

be a connected manifold and

\tilde{B}

its universal cover equipped with the action of the fundamental group

Γ = π_{1} (B)

by deck transformations. Suppose that there is given a homomorphism

φ : Γ \to Diff (F)

Γ

to the group

Diff (F)

of diffeomorphisms of a smooth manifold

F

. Define a manifold

M = \tilde{B} \times_{Γ} F

as the quotient of the manifold

\tilde{B} \times F

by the action of

Γ

given, for any

γ \in Γ

, by

γ (b, f) = (γ b, φ (γ) f), (b, f) \in \tilde{B} \times F .

There is a natural foliation

ℱ

M

, whose leaves are the images of the sets

\tilde{B} \times {f}, f \in F,

under the projection

\tilde{B} \times F \to M

. If, for any

γ \in Γ, γ \neq e

, the diffeomorphism

φ (γ)

has no fixed points, all leaves of

ℱ

are diffeomorphic to

\tilde{B}

There is defined the bundle

π : M \to B : [(b, f)] \mapsto b m o d Γ

such that the leaves of

ℱ

are transverse to the fibers of

π

. The bundle

π : M \to B

is often said to be a flat foliated bundle.

A foliation

ℱ

determines a subbundle

F = T ℱ

of the tangent bundle

T M

, called the tangent bundle to

ℱ

. It consists of all vectors, tangent to the leaves of

ℱ

. Denote by

X (M) = C^{\infty} (M, T M)

the Lie algebra of all smooth vector fields on

M

with the Lie bracket and by

X (ℱ) = C^{\infty} (M, F)

the subspace of vector fields on

M

, tangent to the leaves of

ℱ

at each point. The subspace

X (ℱ)

is a subalgebra of the Lie algebra

X (M)

. Moreover, by the Frobenius theorem, a subbundle

E

T M

is the tangent bundle to a foliation if and only if it is involutive, that is, the space of sections of this bundle is a Lie subalgebra of the Lie algebra

X (M)

: for any

X, Y \in C^{\infty} (M, E)

we have

[X, Y] \in C^{\infty} (M, E)

Let us introduce the following objects:

$∙$ $τ = T M / T ℱ$ is the normal bundle to $ℱ$ ;
$∙$ $P_{τ} : T M \to τ$ is the natural projection;
$∙$ $N^{*} ℱ = {ν \in T^{*} M : 〈 ν, X 〉 = 0 for any X \in F}$ is the conormal bundle to $ℱ$ .

Usually, we will denote by

(x, y) \in I^{p} \times I^{q}

(

I = (0, 1)

is the open interval) the local coordinates in a foliated chart

φ : U \to I^{p} \times I^{q}

and by

(x, y, ξ, η) \in I^{p} \times I^{q} \times R^{p} \times R^{q}

the local coordinates in the corresponding chart on

T^{*} M

Then the subset

N^{*} ℱ \cap π^{- 1} (U) = U_{1}

(here

π : T^{*} M \to M

is the bundle map) is given by the equation

ξ = 0

. Therefore,

φ

determines naturally a foliated chart

φ_{n} : U_{1} \to I^{p} \times I^{q} \times R^{q}

N^{*} ℱ

with the coordinates

(x, y, η)

Any

q

-dimensional distribution

Q \subset T M

such that

T M = F \oplus Q

is called a distribution transverse to the foliation or a connection on the foliated manifold

(M, ℱ)

. Any Riemannian metric

g

M

determines a transverse distribution

H

, which is given by the orthogonal complement of

F

with respect to the given metric:

H = F^{⊥} = {X \in T M : g (X, Y) = 0 for any Y \in F}

Definition 2.8. A vector field

V

on a foliated manifold

(M, ℱ)

is called an infinitesimal transformation of

ℱ

[V, X] \in X (ℱ)

for any

X \in X (ℱ)

The set of all infinitesimal transformations of

ℱ

is denoted by

X (M / ℱ)

V \in X (M / ℱ)

and

T_{t} : M \to M, t \in R

is the flow of the vector field

V

, then the diffeomorphisms

T_{t}

are automorphisms of the foliated manifold

(M, ℱ)

, that is, they take each leaf of

ℱ

to, possibly, another leaf.

Definition 2.9. A vector field

V

on a foliated manifold

(M, ℱ)

is called projectable, if its normal component

P_{τ} (V)

is locally the lift of a vector field on the local base.

In other words, a vector field

V

M

is projectable, if in any foliated chart with local coordinates

(x, y), x \in R^{p}, y \in R^{q},

it has the form

V = \sum_{i = 1}^{p} f^{i} (x, y) \frac{\partial}{\partial x^{i}} + \sum_{j = 1}^{q} g^{j} (y) \frac{\partial}{\partial y^{j}} .

There is a natural action of the Lie algebra

X (ℱ)

C^{\infty} (M, τ)

. The action of

X \in X (ℱ)

N \in C^{\infty} (M, τ)

is given by

θ (X) N = P_{τ} [X, \tilde{N}],

where

\tilde{N} \in X (M)

is any vector field on

M

such that

P_{τ} (\tilde{N}) = N

. A vector field

N \in X (M)

is projectable if and only if its transverse component

P_{τ} (N) \in C^{\infty} (M, τ)

is invariant under the

X (ℱ)

-action

θ

. From here, one can easily see that a vector field on a foliated manifold is called projectable if and only if it is an infinitesimal transformation of the foliation.

2.2 Holonomy

Let

(M, ℱ)

be a foliated manifold. The holonomy map is a generalization to the case of foliations of the first return map (or the Poincaré map) for flows.

Definition 2.10. A smooth transversal is a compact

q

-dimensional manifold

T

, possibly with boundary, and an embedding

i : T \to M

, whose image is everywhere transverse to the leaves of

ℱ

T_{i (t)} i (T) \oplus T_{i (t)} ℱ = T_{i (t)} M

for any

t \in T

We will identify a transversal

T

with the image

i (T) \subset M

Definition 2.11. A transversal is complete, if it meets every leaf of the foliation.

Take an arbitrary continuous leafwise path

γ

with endpoints

γ (0) = x

and

γ (1) = y

. (We will call a path

γ : [0, 1] \to M

leafwise, if its image

γ ([0, 1])

is contained entirely in one leaf of the foliation). Let

T_{0}

and

T_{1}

are smooth transversals such that

x \in T_{0}

and

y \in T_{1}

Choose a partition

t_{0} = 0 < t_{1} < \dots < t_{k} = 1

of the segment

[0.1]

such that for any

i = 1, \dots, k

the curve

γ ([t_{i - 1}, t_{i}])

is contained in some foliated chart

U_{i}

. Shrinking, if necessary, the neighborhoods

U_{1}

and

U_{2}

, one can assume that, for any plaque

P_{1}

U_{1}

, there is a unique plaque

P_{2}

U_{2}

, which meets

P_{1}

. Shrinking, if necessary, the neighborhoods

U_{1}

U_{2}

and

U_{3}

, one can assume that, for any plaque

P_{2}

U_{2}

, there is a unique plaque

P_{3}

U_{3}

, which meets

P_{2}

and so on. After all, we get a family

{U_{1}, U_{2}, \dots, U_{k}}

of foliated coordinate neighborhoods, which covers the curve

γ ([0, 1])

, such that, for any

i = 1, \dots, k

and for any plaque

P_{i - 1}

U_{i - 1}

, there is a unique plaque

P_{i}

U_{i}

, which meets

P_{i - 1}

. In particular, we get a one-to-one correspondence between the plaques of

U_{1}

and the plaques of

U_{k}

The smooth transversal

T_{0}

determines a parametrization of the plaques of

U_{1}

near

x

. Accordingly, a smooth transversal

T_{1}

determines a parametrization of the plaques of

U_{k}

near

y

. Taking into account one-to-one correspondence between the plaques of

U_{1}

and

U_{k}

constructed above, we get a diffeomorphism

H_{T_{0} T_{1}} (γ)

of some neighborhood of

x

T_{0}

to some neighborhood of

y

T_{1}

, which is called the holonomy map along the path

γ

One can easily see that the germ of

H_{T_{0} T_{1}} (γ)

x

does not depend on the choice of a partition

t_{0} = 0 < t_{1} < \dots < t_{k} = 1

of the segment

[0.1]

and of a family of foliated coordinate neighborhoods

{U_{1}, U_{2}, \dots, U_{k}}

. Moreover, the germ of

H_{T_{0} T_{1}} (γ)

x

is not changed, if we replace

γ

by any other continuous leafwise path

γ_{1}

with the initial point

x

and the final point

y

, which is homotopic to

γ

in the class of continuous leafwise paths with the initial point

x

and the final point

y

There is a slightly different definition of holonomy [88] . First, let us introduce some notions.

Definition 2.12. A map

f : V \subset M \to R^{q}

is called distinguished, if in a neighborhood of any point in

V

there exists a foliated chart

(U, φ)

such that the restriction of

f

U

has the form

p r_{n q} \circ φ

, where

p r_{n q} : R^{n} = R^{p} \times R^{q} \to R^{q}

is the natural projection.

Consider the set

D

of germs of distinguished maps at various points of

M

(or, briefly speaking, the set of distinguished germs). Let

σ : D \to M

be the map, which associates to a distinguished germ at

x \in M

the point

x

Let

(U, φ)

be a foliated chart,

P

its plaque. Consider the subset

\tilde{P}

D

, which consists of all germs of the corresponding distinguished map

p r_{n q} \circ φ : U \to R^{q}

in different points of

P

. Sets of the form

\tilde{P}

, determined by the plaques

P

of all possible foliated charts, form a base of a topology on

D

, which endows

D

with the structure of a

p

-dimensional manifold. This topology is called the leaf topology on

D

. It can be easily checked [88] that

σ

is a covering map from the manifold

D

to the manifold

ℱ

Now consider a continuous leafwise path

γ

with the initial point

x

and the final point

y

. Let

π \in σ^{- 1} (x)

, and let

\tilde{γ}

be the lift of

γ

D

with the initial point

π

via the covering map

σ

. The holonomy map associated with

γ

is the map

h_{γ} : σ^{- 1} (x) \to σ^{- 1} (y),

which takes

π \in σ^{- 1} (x)

to the final point of the path

\tilde{γ}

The connection between the holonomy maps defined above is established as follows. Let

T_{0}

and

T_{1}

be smooth transversals such that

x \in T_{0}

and

y \in T_{1}

. Let

π \in σ^{- 1} (x)

, and let

f : U \subset M \to R^{q}

be a distinguished map defined in a neighborhood of

x

. A choice of a distinguished map

f

defined in a neighborhood of

x

is equivalent to a choice of a local coordinate system

f \circ i

T_{0}

defined in a neighborhood of

x

. The diffeomorphism

H_{T_{0} T_{1}} (γ)

allows to define a local coordinate system on

T_{1}

defined in some neighborhood of

y

, that, in its turn, gives a distinguished map

f_{1} : V \subset M \to R^{q}

defined in a neighborhood of

y

. The germ of the distinguished map

f_{1}

y

coincides with

h_{γ} (π) \in σ^{- 1} (y)

γ

is a closed leafwise path with the initial and final points

x

, and

T

is a smooth transversal such that

x \in T

, then

H_{T T} (γ)

is a local diffeomorphism of

T

, which leaves

x

fixed. The correspondence

γ \to H_{T T} (γ)

defines a group homomorphism

H_{T} : π_{1} (L_{x}, x) \to {Diff}_{x} (T)

from the fundamental group

π_{1} (L_{x}, x)

of the leaf

L_{x}

to the group

{Diff}_{x} (T)

of germs at

x

of local diffeomorphisms of

T

, which leave

x

fixed. The image of the homomorphism

H_{T}

is called the holonomy group of the leaf

L_{x}

x

The holonomy group of a leaf

L

at a point

x \in L

is independent modulo isomorphism of the choice of a transversal

T

and

x

. A leaf is said to have trivial holonomy, if its holonomy group is trivial.

Example 2.13. Let

X

be a complete nonsingular vector field on a manifold

M

of dimension

n

x_{0}

a (for simplicity, isolated) periodic point of the flow

X_{t}

of the given vector field and

C

the corresponding closed phase curve. Let

T

be an

(n - 1)

-dimensional submanifold of

M

, passing through

x_{0}

transverse to the vector

X (x_{0})

T_{x} M = T_{x} T \oplus R X (x_{0})

. For all

x \in T

, closed enough to

x_{0}

, there is the least

t (x) > 0

such that the corresponding positive semi-trajectory of the flow

{X_{t} (x) : t > 0}

meets

T

X_{t (x)} (x) \in T

Thus, we get a local diffeomorphism

φ_{T} : x \mapsto X_{t (x)} (x)

T

, defined in a neighborhood of

x_{0}

and taking

x_{0}

to itself. This diffeomorphism is called the first return map (or the Poincaré map) along the curve

C

ℱ

is the foliation on

M

given by the trajectories of

X

, then the holonomy group of the leaf

C

coincides with

Z

, and the germ of

φ_{T}

x_{0}

is a generator of this group.

Example 2.14. For the Reeb foliation of the three-dimensional sphere

S^{3}

all noncompact leaves have trivial holonomy. The holonomy group of the compact leaf is isomorphic to

Z^{2}

For any smooth transversal

T

and for any

x \in T

, there is a natural isomorphism of the tangent space

T_{x} T

with the normal space

τ_{x}

ℱ

. Thus, the normal bundle

τ

plays a role of the tangent bundle to the (germs of ) transversals to

ℱ

. For any continuous leafwise path

γ

with the initial point

x

and the final point

y

and for any smooth transversals

T_{0}

and

T_{1}

such that

x \in T_{0}

and

y \in T_{1}

, the differential of the holonomy map

H_{T_{0} T_{1}} (γ)

x

defines a linear map

d H_{T_{0} T_{1}} (γ)_{x} : τ_{x} \to τ_{y}

. It is easy to check that this map is independent of the choice of transversals

T_{0}

and

T_{1}

. It is called the linear holonomy map and is denoted by

d h_{γ} : τ_{x} \to τ_{y}

. Taking the adjoint of

d h_{γ}

, one gets a linear map

d h_{γ}^{*} : N^{*} ℱ_{y} \to N^{*} ℱ_{x}

Now we turn to another notion related with the holonomy, the notion of holonomy pseudogroup. First, recall the general definition of a pseudogroup.

Definition 2.15. A family

Γ

, consisting of diffeomorphisms between open subsets of a manifold

X

(or, in other words, of local diffeomorphisms of

X

) is called a pseudogroup on

X

, if the following conditions hold:

(1) if $Φ \in Γ$ , then $Φ^{- 1} \in Γ$ ;
(2) if $Φ_{1} : U \to U_{1}$ and $Φ_{2} : U_{1} \to U_{2}$ belong to $Γ$ , then $Φ_{2} \circ Φ_{1} : U \to U_{2}$ belongs to $Γ$ ;
(3) if $Φ : U \to U_{1}$ belongs to $Γ$ , then its restriction to any open subset $V \subset U$ belongs to $Γ$ ;
(4) if a diffeomorphism $Φ : U \to U_{1}$ coincides on some neighborhood of each point in $U$ with an element of $Γ$ , then $Φ \in Γ$ ;
(5) the identity diffeomorphism belongs to $Γ$ .

Example 2.16. The set of all local diffeomorphisms of a manifold

X

form a pseudogroup on

X

. One can also consider pseudogroups, consisting of local diffeomorphisms of a manifold

X

, which preserve a geometric structure, for instance, the pseudogroup of local isometries and so on.

Definition 2.17. Let

(M, ℱ)

be a smooth foliated manifold and

X

the disjoint union of all smooth transversals to

ℱ

. The holonomy pseudogroup of the foliation

ℱ

is the pseudogroup

Γ

, which consists of all local diffeomorphisms of

X

, whose germ at any point coincides with the germ of the holonomy map along a leafwise path.

Definition 2.18. Let

(M, ℱ)

be a smooth foliated manifold and

T

be a smooth transversal. The holonomy pseudogroup induced by the foliation

ℱ

T

is the pseudogroup

Γ_{T}

, which consists of all local diffeomorphisms of

T

, whose germ at any point coincides with the germ of the holonomy map along a leafwise path.

There is a special class of smooth transversals given by good covers of the manifold

M

Definition 2.19. A foliated chart

φ : U \subset M \to R^{p} \times R^{q}

is called regular, if it admits an extension to a foliated chart

\bar{φ} : V \to R^{p} \times R^{q}

such that

\bar{U} \subset V

Definition 2.20. A cover of a manifold

M

by foliated neighborhoods

{U_{i}}

is called good, if:

(1) Any chart $(U_{i}, φ_{i})$ is a regular foliated chart;
(2) If ${\bar{U}}_{i} \cap {\bar{U}}_{j} \neq \emptyset$ , then $U_{i} \cap U_{j} \neq \emptyset$ and the set $U_{i} \cap U_{j}$ is connected. The same is true for the corresponding foliated neighborhoods $V_{i}$ ;
(3) Each plaque of $V_{i}$ meets at most one plaque of $V_{j}$ . A plaque of $U_{i}$ meets a plaque of $U_{j}$ if and only if the intersection of the corresponding plaques of $V_{i}$ and $V_{j}$ is nonempty.

Good covers always exist.

Let

U = {U_{i}}

be a good cover for the foliation

ℱ

φ_{i} : U_{i} \to \sim = I^{p} \times I^{q}

. For any

i

, put

T_{i} = φ_{i}^{- 1} ({0} \times I^{q}) .

Then

T_{i}

is a transversal, and

T = \cup T_{i}

is a complete transversal. For

y \in T_{i}

denote by

P_{i} (y)

the plaque of

U_{i}

, passing through

y

. For any pair of indices

i

and

j

such that

U_{i} \cap U_{j} \neq \emptyset

, define

T_{i j} = {y \in T_{i} : P_{i} (y) \cap U_{j} \neq \emptyset} .

There is defined a transition function

f_{i j} : T_{i j} \to T_{j i}

given for

y \in T_{i j}

by the formula

f_{i j} (y) = y_{1},

where

y_{1} \in T_{j i}

corresponds to the unique plaque

P_{j} (y_{1})

, for which

P_{i} (y) \cap P_{j} (y_{1}) \neq \emptyset .

The holonomy pseudogroup

Γ_{T}

induced by

ℱ

T

coincides with the pseudogroup generated by the maps

f_{i j}

Definition 2.21 (cf., for instance, [92] ). A transverse structure on a foliation

ℱ

is a structure on a complete transversal

T

, invariant under the action of the holonomy pseudogroup

Γ_{T}

Using the notion of transverse structure, one can single out classes of foliations with specific transverse properties. For instance, if a complete transversal

T

is equipped with a Riemannian metric, and the holonomy pseudogroup

Γ_{T}

consists of all local isometries of this Riemannian metric, we get a class of Riemannian foliations (see Section 2.5 ). Similarly, if a complete transversal

T

is equipped with a symplectic structure, and the holonomy pseudogroup

Γ_{T}

consists of all local diffeomorphisms, preserving this symplectic structure, we get a class of symplectic foliations (see Section 2.6 ).

One can also consider Kaehler foliations, measurable foliations and so on.

2.3 Transverse measures

Before we turn to the discussion of an analogue of the notion of measure on the leaf space of a foliation, we recall some basic facts, concerning to densities and integration of densities (cf., for instance, [31, 86] ).

Definition 2.22. Let

L

be an

n

-dimensional linear space and

ℬ (L)

the set of bases in

L

. An

α

-density on

L

(

α \in R

) is a function

ρ : ℬ (L) \to C

such that, for any

A = (A_{i j}) \in G L (n, C)

and

e = (e_{1}, e_{2}, \dots, e_{n}) \in ℬ (L)

ρ (e \cdot A) = | det A |^{α} ρ (e),

where

(e \cdot A)_{i} = \sum_{j = 1}^{n} e_{j} A_{j i}, i = 1, 2, \dots, n

We will denote by

| L |^{α}

the space of all

α

-densities on

L

. For any vector bundle

V

M

, denote by

| V |^{α}

the associated bundle of

α

-densities,

| V | = | V |^{1}

For any smooth, compactly supported density

ρ

on a smooth manifold

M

there is a well-defined integral

\int_{M} ρ

, independent of the fact if

M

is orientable or not. This fact allows to define a Hilbert space

L^{2} (M)

, canonically associated with

M

, which consists of square integrable half-densities on

M

. The diffeomorphism group of

M

acts on the space

L^{2} (M)

by unitary transformations.

Definition 2.23. A (Borel) transversal to a foliation

ℱ

is a Borel subset of

M

, which intersects each leaf of the foliation in an at most countable set.

Definition 2.24. A transverse measure

Λ

is a countably additive Radon measure, defined on the set of all transversals to the foliation.

Definition 2.25. A transverse measure

Λ

is called holonomy invariant, if for any transversals

B_{1}

and

B_{2}

and any bijective Borel map

φ : B_{1} \to B_{2}

such that, for any

x \in B_{1}

, the point

φ (x)

belongs to the leaf through the point

x

, we have:

Λ (B_{1}) = Λ (B_{2})

Example 2.26. Let us call by a transverse density any section of the bundle

| τ |

. Since, for any smooth transversal

T

, there is a canonical isomorphism

T_{x} T \sim = τ_{x}

, a continuous positive density

ρ \in C (M, | τ |)

determines a continuous positive density on

T

, that determines a transverse measure. This transverse measure is holonomy invariant if and only if

ρ

is invariant under the linear holonomy action.

Example 2.27. Any compact leaf

L

of the foliation

ℱ

defines a holonomy invariant transverse measure

Λ

. For any transversal

T

and for any set

A \subset T

, its measure

Λ (A)

equals the number of elements in

A \cap L

Let

α \in C^{\infty} (M, | T ℱ |)

be a smooth positive leafwise density on

M

Starting from a transverse measure

Λ

and the density

α

, one can construct a Borel measure

μ

M

in the following way. Take a good cover

{U_{i}}

M

by foliated coordinate neighborhoods with the corresponding coordinate maps

φ_{i} : U_{i} \to I^{p} \times I^{q}

and a partition of unity

{ψ_{i}}

subordinate to this cover. Consider the corresponding complete transversal

T = \cup_{i} T_{i}

, where

T_{i} = φ_{i}^{- 1} ({0} \times I^{q}) .

In any foliated chart

(U_{i}, φ_{i})

, the transverse measure

Λ

defines a measure

Λ_{i}

T_{i}

, and the smooth positive leafwise density

α

defines a family

{α_{i, y} : y \in T_{i}}

, where

{α_{i, y}}

is a smooth positive density on the plaque

P_{i} (y)

. Observe that

Λ

is holonomy invariant if and only if for any pair of indices

i

and

j

such that

U_{i} \cap U_{j} \neq \emptyset

, we have:

f_{i j} (Λ_{i}) = Λ_{j}

For any

u \in C_{c}^{\infty} (M)

, put

\int_{M} u (m) d μ (m) = \sum_{i} \int_{T_{i}} \int_{P_{i} (y)} ψ_{i} (x, y) u (x, y) d α_{i, y} (x) d Λ_{i} (y) .

One can show that this formula defines a measure

μ

M

, which is independent of the choice of a cover

{U_{i}}

and a partition of unity

{ψ_{i}}

A measure

μ

M

will be called holonomy invariant, if it is obtained by means of the above construction from a holonomy invariant transverse measure

Λ

with some choice of a smooth positive leafwise density

α

If we take in the above construction instead of the leafwise density

u \cdot α

the restrictions to the leaves of an arbitrary differential

p

-form

ω

M

, we obtain a well-defined functional

C

C_{c}^{\infty} (M, \land^{p} T^{*} M)

, called the Ruelle-Sullivan current [161] , corresponding to

Λ

〈 C, ω 〉 = \sum_{i} \int_{T_{i}} \int_{P_{i} (y)} ψ_{i} (x, y) ω_{i, y} (x) d Λ_{i} (y), ω \in C_{c}^{\infty} (M,^{p} ⋀ T^{*} M),

where

ω_{i, y}

is the restriction of

ω

to the plaque

P_{i} (y), y \in T_{i}

A transverse measure

Λ

is holonomy invariant, if and only if the corresponding Ruelle-Sullivan current

C

is closed:

〈 C, d σ 〉 = 0, σ \in C_{c}^{\infty} (M,^{p - 1} ⋀ T^{*} M) .

Example 2.28. Suppose that a transverse measure

Λ

is given by a smooth positive transverse density

ρ \in C^{\infty} (M, | τ |)

. Take a positive leafwise density

α \in C^{\infty} (M, | T ℱ |)

. Then the corresponding measure

μ

M

is given by the smooth positive density

α \otimes ρ \in C^{\infty} (M, | T M |)

, which corresponds to

α

and

ρ

under the canonical isomorphism

| T M | \sim = | T ℱ | \otimes | τ |

, defined by the short exact sequence

0 \to T ℱ \to T M \to τ \to 0

Example 2.29. Suppose that a holonomy invariant transverse measure

Λ

is given by a compact leaf

L

of the foliation

ℱ

, and

α \in C^{\infty} (M, | T ℱ |)

is a smooth positive leafwise density on

M

. Then the corresponding measure

μ

M

is the

δ

-function along

L

\int_{M} f (x) d μ (x) = \int_{L} f (x) α (x), f \in C_{c} (M) .

Example 2.30. Suppose that the foliation

ℱ

is given by the orbits of a locally free action of a Lie group

H

on the compact manifold

M

and a smooth leafwise density

α

is given by a fixed Haar measure

d h

H

. Then the corresponding measure

μ

M

is holonomy invariant if and only if it is invariant under the action of

H

2.4 Connections

The infinitesimal expression of the holonomy on a foliated manifold is the canonical flat connection

\overset{\circ}{\nabla} : X (ℱ) \times C^{\infty} (M, τ) \to C^{\infty} (M, τ)

in the normal bundle

τ

, defined along the leaves of

ℱ

(the Bott connection) [22, 23] . It is given by

\begin{matrix} {\overset{\circ}{\nabla}}_{X} N = θ (X) N = P_{τ} [X, \tilde{N}], X \in X (ℱ), N \in C^{\infty} (M, τ), \end{matrix}

(2.1)

where

\tilde{N} \in C^{\infty} (M, T M)

is any vector field on

M

such that

P_{τ} (\tilde{N}) = N

Thus, the restriction of

τ

to any leaf of

ℱ

is a flat vector bundle. The parallel transport in

τ

along any leafwise path

γ : x \to y

defined by

\overset{\circ}{\nabla}

coincides with the linear holonomy map

d h_{γ} : τ_{x} \to τ_{y}

Definition 2.31. A connection

\nabla : X (M) \times C^{\infty} (M, τ) \to C^{\infty} (M, τ)

in the normal bundle

τ

is called adapted, if its restriction to

X (ℱ)

coincides with the Bott connection

\overset{\circ}{\nabla}

One can construct an adapted connection, starting with an arbitrary Riemannian metric

g_{M}

M

. Denote by

\nabla^{g}

the Levi-Civita connection, defined by

g_{M}

. An adapted connection

\nabla

is given by

\begin{matrix} \begin{matrix} \nabla_{X} N & = P_{τ} [X, \tilde{N}], X \in X (ℱ), N \in C^{\infty} (M, τ) \\ \nabla_{X} N & = P_{τ} \nabla_{X}^{g} \tilde{N}, X \in C^{\infty} (M, F^{⊥}), N \in C^{\infty} (M, τ), \end{matrix} \end{matrix}

(2.2)

where

\tilde{N} \in C^{\infty} (M, T M)

is any vector field such that

P_{τ} (\tilde{N}) = N

. One can show that the adapted connection

\nabla

described above has zero torsion.

The Bott connection

\overset{\circ}{\nabla}

τ

determines a connection

(\overset{\circ}{\nabla})^{*}

τ^{*} \sim = N^{*} ℱ

by the formula

\begin{matrix} (\overset{\circ}{\nabla})_{X}^{*} ω (N) = X [ω (N)] - ω (\nabla_{X} N), \end{matrix}

(2.3)

for any

X \in T M, ω \in C^{\infty} (M, τ^{*}), N \in C^{\infty} (M, τ)

Definition 2.32. An adapted connection

\nabla

in the normal bundle

τ

is called holonomy invariant, if, for any

X \in X (ℱ)

Y \in X (M)

and

N \in C^{\infty} (M, τ)

, we have

(θ (X) \nabla)_{Y} N = 0,

where, by definition,

(θ (X) \nabla)_{Y} N = θ (X) [\nabla_{Y} N] - \nabla_{θ (X) Y} N - \nabla_{Y} [θ_{Y} N] .

A holonomy invariant adapted connection in

τ

is called a basic (or projectable) connection.

A fundamental property of basic connections is the fact that their curvature

R_{\nabla}

is a basic form, i.e.

i_{X} R_{\nabla} = 0, θ (X) R_{\nabla} = 0

for any

X \in X (ℱ)

There are topological obstructions for the existence of basic connections for an arbitrary foliations found independently by Kamber and Tondeur and Molino (cf., for instance, [106, 134] ).

2.5 Riemannian foliations

Definition 2.33. A foliation

(M, ℱ)

is called Riemannian, if it has a transverse Riemannian structure. In other words, a foliation

(M, ℱ)

is called Riemannian, if there is a cover

{U_{i}}

M

by foliated coordinate charts,

φ_{i} : U_{i} \to I^{p} \times I^{q}

, and Riemannian metrics

g^{(i)} (y) = \sum_{α β} g_{α β}^{(i)} (y) d y^{α} d y^{β}

, defined on the local bases

I^{q}

ℱ

such that, for any coordinate transformation

φ_{i j} (x, y) = (α_{i j} (x, y), γ_{i j} (y)), (x, y) \in φ_{j} (U_{i} \cap U_{j}),

the map

γ_{i j}

preserves the metric on

I^{q}

γ_{i j}^{*} (g^{(j)}) = g^{(i)}

The class of Riemannian foliations was introduced in the papers of Reinhart [154, 155] . There are several equivalent characterizations of Riemannian foliations.

Before we formulate the corresponding result (cf., for instance, [153,ChapterIV] ), we introduce some auxiliary notions.

Definition 2.34. A distribution on a Riemannian manifold is called totally geodesic, if every geodesic, which is tangent to the given distribution at some point, is tangent to it along the whole its length.

Definition 2.35. The second fundamental form of a distribution

H

on a Riemannian manifold

(M, g)

is the tensor

S

(which takes any vector

X \in T_{x} M

at a point

x \in M

to a linear map

S_{X} : T_{x} M \to T_{x} M

) given by

g (S_{M} N, X) = \frac{1}{2} g (\nabla_{M}^{g} N + \nabla_{N}^{g} M, X), g (S_{M} N, L) = 0, g (S_{M} X, Y) = 0, g (S_{M} X, N) = g (S_{M} N, X), S_{X} = 0,

where

X, Y \in X (ℱ)

and

L, M, N \in C^{\infty} (M, F^{⊥})

\nabla^{g}

denote the Levi-Civita connection determined by

g

Definition 2.36. A map

f : M \to B

of Riemannian manifolds

M

and

B

is called a Riemannian submersion, if the tangent map

d f_{m} : T_{m} M \to T_{f (m)} B

at any point

m \in M

is surjective and induces an isometric map between the normal space

T_{m} M / T_{m} f^{- 1} (f (m))

to the level set

f^{- 1} (f (m))

f

m

and the tangent space

T_{f (m)} B

B

f (m)

Theorem 2.37. A foliation

(M, ℱ)

is Riemannian if and only if there is a Riemannian metric

g

M

, satisfying any of the following equivalent conditions:

(1) The distribution $H = F^{⊥}$ is totally geodesic.
(2) The second fundamental form of $H$ vanishes.
(3) The induced metric $g_{τ}$ on the normal bundle $τ$ is holonomy invariant: ${\overset{\circ}{\nabla}}_{X} g_{τ} (M, N) = 0$ for any $X \in X (ℱ)$ and for any $M, N \in C^{\infty} (M, τ)$ , where, by definition, ${\overset{\circ}{\nabla}}_{X} g_{τ} (M, N) = X [g_{τ} (M, N)] - g_{τ} ({\overset{\circ}{\nabla}}_{X} M, N) - g_{τ} (M, {\overset{\circ}{\nabla}}_{X} N) .$
(4) For any vector fields $M$ and $N$ , which are defined on an open set, are orthogonal to the leaves and are infinitesimal transformations of the foliation, and for any $X \in X (ℱ)$ , we have $X [g (M, N)] = 0 .$
(5) In any foliated chart $φ : U \to I^{p} \times I^{q}$ with the local coordinates $(x, y)$ , the restriction $g_{H}$ of $g$ to $H$ is written in the form $g_{H} = \sum_{α, β = 1}^{q} g_{α β} (y) θ^{α} θ^{β},$ where $θ^{α} \in H^{*}$ is the 1-form, corresponding to the form $d y^{α}$ under the isomorphism $H^{*} \overset{\sim =}{\to} T^{*} R^{q}$ , and $g_{α β} (y)$ depend only on the transverse variables $y \in R^{q}$ .
(6) $(M, ℱ)$ locally has the structure of a Riemannian submersion, i. e., for any foliated chart $φ : U \to I^{p} \times I^{q}$ , there exists a Riemannian metric on $I^{q}$ such that the corresponding distinguished map $p r_{n q} \circ φ : U \to I^{q}$ is a Riemannian submersion.
(7) The adapted connection $\nabla$ on the normal bundle $τ$ given by ( 2.2 ) is a (torsion-free) Riemannian connection: for any $Y \in X (M)$ and $M, N \in C^{\infty} (M, τ)$ $Y [g_{τ} (M, N)] = g_{τ} (\nabla_{Y} M, N) + g_{τ} (M, \nabla_{Y} N) .$
(8) The holonomy group of the adapted connection $\nabla$ on the normal bundle $τ$ given by ( 2.2 ) at any point preserves the metric.

Definition 2.38. Any Riemannian metric on

M

, satisfying the equivalent conditions of Theorem 2.37 , is called bundle-like.

One can prove that a (torsion-free) Riemannian connection on the normal bundle

τ

to a Riemannian foliation

ℱ

is unique. It is uniquely determined by the transverse metric

g_{τ}

and is called the transverse Levi-Civita connection for

ℱ

. Thus, the transverse Levi-Civita connection is an adapted connection.

Moreover, the transverse Levi-Civita connection turns out to be a holonomy invariant and, therefore, a basic connection. In particular, this shows the existence of a basic connection for any Riemannian foliation.

The existence of a bundle-like metric on a foliated manifold imposes strong restrictions on the geometry of the foliation. There are structure theorems for Riemannian foliations obtained by Molino (cf. [136, 135] ). Using these structure theorems, many questions, concerning to Riemannian foliations, can be reduced to the case of Lie foliations, that is, of foliations, whose transverse structure is modeled by a finite-dimensional Lie group (cf. Example 2.41 ).

Example 2.39. Any foliation defined by a submersion

π : M \to B

is Riemannian.

Example 2.40. The orbits of a locally free isometric action of a Lie group on a Riemannian manifold define a Riemannian foliation.

On the other hand, flows, whose orbits form a Riemannian foliation, are called Riemannian flows. There are examples of Riemannian flows, which are not isometric. Concerning to Riemannian flows, see, for instance, [32] , and also [137,AppendixA] .

Example 2.41. Let

M

be a smooth manifold,

g

a real finite-dimensional Lie algebra and

ω

an 1-form on

M

with values in

g

, satisfying the conditions:

(1) the map $ω_{x} : T_{x} M \to g$ is surjective for any $x \in M$ ;
(2) $d ω + \frac{1}{2} [ω, ω] = 0$ .

The distribution

F_{x} = ker ω_{x}

is integrable and, therefore, defines a codimension

q = dim g

foliation on

M

. Such a foliation is called a Lie

g

-foliation. The class of Lie foliations was introduced in [69] .

Any Lie foliation is Riemannian. In the case

g = R

, a Lie foliation is precisely a codimension one foliation given by a non-vanishing closed 1-form. Actually, it can be easily seen that a codimension one foliation is Riemannian if and only if it is given by a non-vanishing closed 1-form.

Example 2.42. A foliation

ℱ

on a manifold

M = \tilde{B} \times_{Γ} F

, obtained by the suspension from a manifold

B

and a homomorphism

φ : Γ \to Diff (F)

of the fundamental group

Γ = π_{1} (B)

is Riemannian if and only if for any

γ \in Γ

the diffeomorphism

φ (γ)

preserves a Riemannian metric on

F

Let

ℱ

be a transversely oriented Riemannian foliation and

g

a bundle-like Riemannian metric. The induced metric on the normal bundle

τ

yields the transverse volume form

v_{τ} \in C^{\infty} (M, \land^{q} τ^{*}) = C^{\infty} (M, \land^{q} N^{*} ℱ)

, which is holonomy invariant and defines, therefore, a holonomy invariant transverse measure on

ℱ

2.6 Symplectic foliations

Definition 2.43. A foliation

(M, ℱ)

is called symplectic, if it has a transverse symplectic structure. In other words, a foliation

(M, ℱ)

is called symplectic, if there is a cover

{U_{i}}

M

by foliated coordinate charts,

φ_{i} : U_{i} \to I^{p} \times I^{q}

, and symplectic forms

ω_{i}

, defined on the local bases

I^{q}

ℱ

such that, for any coordinate transformation

φ_{i j} (x, y) = (α_{i j} (x, y), γ_{i j} (y)), (x, y) \in φ_{j} (U_{i} \cap U_{j}),

the map

γ_{i j}

preserves the symplectic form,

γ_{i j}^{*} ω_{j} = ω_{i}

Definition 2.44. A presymplectic manifold is a manifold equipped with a closed

2

-form of constant rank.

A transverse symplectic structure on

(M, ℱ)

uniquely determines a presymplectic structure

ω

M

such that

T ℱ

coincides with the kernel of

ω

. On the other hand, if

(M, ω)

is a presymplectic manifold, then the kernel of

ω

determines an integrable distribution on

M

and

ω

induces a transverse symplectic structure on the corresponding foliation

(M, ℱ)

(cf. for instance, [16, 20] , in [16] symplectic foliations are called Hamiltonian). If

(M, ℱ)

is a symplectic foliation and

ω

is the corresponding presymplectic structure on

M

, then the

q

-form

\land^{q} ω

defines the holonomy invariant transverse density

| \land^{q} ω | \in | \land^{q} τ^{*} |

, and, therefore, a holonomy invariant transverse measure, which can be naturally called the transverse Liouville measure.

Example 2.45. A foliation

ℱ

on a manifold

M = \tilde{B} \times_{Γ} F

, obtained by the suspension from a manifold

B

and a homomorphism

φ : Γ \to Diff (F)

of the fundamental group

Γ = π_{1} (B)

is symplectic if and only if for any

γ \in Γ

the diffeomorphism

φ (γ)

preserves a symplectic structure on

F

Example 2.46. Recall (cf., for instance, [120] ) that a submanifold

Σ

of a symplectic manifold

X

is called coisotropic, if, for any

σ \in Σ

, the skew-orthogonal complement

(T_{σ} Σ)^{⊥}

T_{σ} Σ

is contained in

T_{σ} Σ

. If

Σ

is a coisotropic submanifold, then the distribution

(T_{σ} Σ)^{⊥}

is integrable, and the corresponding foliation

ℱ_{Σ}

is called the characteristic foliation of the coisotropic submanifold

Σ

. It is well-known that there is a canonical symplectic structure on

T_{σ} Σ / (T_{σ} Σ)^{⊥}

, therefore, the foliation

ℱ_{Σ}

is symplectic. Moreover, if

ℱ_{Σ}

is simple, then the set

Γ \subset T^{*} M \times T^{*} M

, which consists of all

(ν, ν^{'}) \in T^{*} M \times T^{*} M

such that

ν

and

ν^{'}

lie on the same leaf of

ℱ_{Σ}

is a canonical relation.

As shown in [20] (see also [80] ), any presymplectic manifold can be obtained, using this construction, that is, as a coisotropic submanifold of a symplectic manifold. A particular example of the construction described above is the following one. For a foliated manifold

(M, ℱ)

, consider

T^{*} M

as a symplectic manifold with the standard symplectic structure.

Then

N^{*} ℱ

is a coisotropic submanifold in

T^{*} M

. The corresponding characteristic foliation

ℱ_{N}

is the natural lift of

ℱ

to the conormal bundle and is called a horizontal (or linearized) foliation. Thus, the linearized foliation

ℱ_{N}

is symplectic. The coordinate chart

φ_{n} : N^{*} ℱ \to I^{p} \times I^{q} \times R^{q}

determined by a foliated coordinate chart

φ

M

(cf. Section 2.1 ) is a foliated chart for

ℱ_{N}

with plaques given by the level sets

y = c o n s t, η = c o n s t

. Informally speaking, the leaf space

N^{*} ℱ / ℱ_{N}

ℱ_{N}

can be considered as the cotangent bundle to the leaf space

M / ℱ

ℱ

2.7 Differential operators

Let

(M, ℱ)

be a compact foliated manifold and

E

a smooth vector bundle on

M

(unless otherwise is stated, we will assume that vector bundles under consideration are smooth and complex). We start with some general definitions, concerning to differential operators on

M

Definition 2.47. A linear differential operator

A

of order

μ

, acting in

C^{\infty} (M, E)

, is called a tangential differential operator, if, in any foliated chart

φ : U \subset M \to I^{p} \times I^{q}

and any trivialization of

E

over it,

A

is of the form

\begin{matrix} A = \sum_{| α | \leq μ} a_{α} (x, y) D_{x}^{α}, (x, y) \in I^{p} \times I^{q}, \end{matrix}

(2.4)

where

a_{α}

are matrix-valued function on

I^{p} \times I^{q}

D_{x} = \frac{1}{i} \frac{\partial}{\partial x}

Definition 2.48. For a tangential differential operator

A

given by ( 2.4 ) in some foliated chart

φ : U \subset M \to I^{p} \times I^{q}

and a trivialization of

E

over it, define its tangential (complete) symbol

σ (x, y, ξ) = \sum_{| α | \leq μ} a_{α} (x, y) ξ^{α}, (x, y) \in I^{p} \times I^{q}, ξ \in R^{p},

and its tangential principal symbol

σ_{μ} (x, y, ξ) = \sum_{| α | = μ} a_{α} (x, y) ξ^{α}, (x, y) \in I^{p} \times I^{q}, ξ \in R^{p} .

The tangential principal symbol is invariantly defined as a section of the bundle

ℒ (π_{F}^{*} E)

T^{*} ℱ

(where

π_{F} : T^{*} ℱ \to M

is the natural projection).

Definition 2.49. A tangential differential operator

A

is called tangentially elliptic, if its tangential principal symbol

σ_{μ}

is invertible for

ξ \neq 0

Let

D^{m} (M, E)

denote the set of all differential operators of order

m

and

D^{μ} (ℱ, E)

denote the set of all tangential differential operators of order

μ

, acting in

C^{\infty} (M, E)

Introduce classes

D^{m, μ} (M, ℱ, E)

, which are linearly generated by arbitrary compositions of tangential differential operators of order

μ

and differential operators of order

m

M

. In other words, an operator

A \in D^{m, μ} (M, ℱ, E)

is of the form

A = \sum_{α} B_{α} C_{α}

, where

B_{α} \in D^{m} (M, E)

C_{α} \in D^{μ} (ℱ, E)

. If

A_{1} \in D^{m_{1}, μ_{1}} (M, ℱ, E)

A_{2} \in D^{m_{2}, μ_{2}} (M, ℱ, E)

, then

A_{1} \circ A_{2} \in D^{m_{1} + m_{2}, μ_{1} + μ_{2}} (M, ℱ, E)

and, if

A \in D^{m, μ} (M, ℱ, E)

, then

A^{*} \in D^{m, μ} (M, ℱ, E)

. The classes

D^{m, μ} (M, ℱ, E)

can be extended to classes

Ψ^{m, μ} (M, ℱ, E)

, which contain, for instance, parametrices for elliptic operators of class

D^{m, μ} (M, ℱ, E)

(cf. [113] ).

We will use the standard classes of pseudodifferential operators

Ψ^{k} (M, E)

(for the theory of pseudodifferential operators see, for instance, [101, 175, 179, 167] ). Recall that a pseudodifferential operator on

M

is a linear operator

P : C^{\infty} (M) \to D^{'} (M)

, which can be represented in a coordinate patch

X \subset R^{n}

P u (x) = \int e^{(x - y) ξ} p (x, ξ) u (y) d y d ξ, x \in X,

where

u \in C_{c}^{\infty} (X)

p (x, ξ) \in S^{m} (X \times R^{n})

is the complete symbol of

P

Usually, we will assume that the complete symbol

p

can be represented as an asymptotic sum

p \sim p_{m} + p_{m - 1} + \dots

, where

p_{k}

is homogeneous of degree

k

ξ

for

| ξ | > 1

. The principal symbol

p_{m}

P

is well-defined as a section of the bundle

ℒ (π^{*} E)

{\tilde{T}}^{*} M = T^{*} M \ {0}

, where

π : T^{*} M \to M

is the natural projection.

Definition 2.50. The transversal principal symbol

σ_{P}

of an operator

P \in Ψ^{m} (M, E)

is the restriction of its principal symbol

p_{m}

{\tilde{N}}^{*} ℱ = N^{*} ℱ \ {0}

Definition 2.51. A operator

P \in Ψ^{m} (M, E)

is called transversally elliptic, if its transversal principal symbol

σ_{P} (ν)

is invertible for any

ν \in {\tilde{N}}^{*} ℱ

Now suppose that a closed foliated manifold

(M, ℱ)

is equipped with a Riemannian metric

g_{M}

. Let

H

be the orthogonal complement of

F = T ℱ

Thus, there is a decomposition of

T M

into the direct sum

T M = F \oplus H

and the corresponding bigrading of the exterior power bundle

Λ^{*} T^{*} M

Λ^{k} T^{*} M = \oplus_{i + j = k} Λ^{i, j} T^{*} M, Λ^{i, j} T^{*} M = Λ^{i} H^{*} \otimes Λ^{j} F^{*} .

There is (cf., for instance, [15,Proposition10.1] ,[177] ) the corresponding decomposition of the de Rham differential

d

into the sum of bigraded components of the form

\begin{matrix} d = d_{F} + d_{H} + θ . \end{matrix}

(2.5)

Here

(1) $d_{F} = d_{0, 1} : C^{\infty} (M, Λ^{i, j} T^{*} M) \to C^{\infty} (M, Λ^{i, j + 1} T^{*} M)$ is the tangential de Rham differential, which is a first order tangentially elliptic operator, independent of the choice of $g$ ;
(2) $d_{H} = d_{1, 0} : C^{\infty} (M, Λ^{i, j} T^{*} M) \to C^{\infty} (M, Λ^{i + 1, j} T^{*} M)$ is the transversal de Rham differential, which is a first order transversally elliptic operator;
(3) $θ = d_{2, - 1} : C^{\infty} (M, Λ^{i, j} T^{*} M) \to C^{\infty} (M, Λ^{i + 2, j - 1} T^{*} M)$ is a zero order differential operator, which is the contraction operator by the 2-form $θ$ on $M$ with values in $F$ , $θ \in C^{\infty} (M, F \otimes Λ^{2} τ^{*})$ , given by $θ (X, Y) = p_{F} ([X, Y]), X, Y \in C^{\infty} (M, H),$ where $P_{F} : T M \to F$ is the natural projection. In particular, $θ$ vanishes if and only if $H$ is integrable.

There is a similar decomposition for the adjoint:

\begin{matrix} δ = δ_{F} + δ_{H} + θ^{*}, \end{matrix}

(2.6)

where

δ_{F}

δ_{H}

and

θ^{*}

are the adjoints of

d_{F}

d_{H}

and

θ

in the Hilbert space

L^{2} (M, Λ T^{*} M)

accordingly.

The Laplace operator

Δ_{g} = d δ + δ d

of the metric

g_{M}

can be written in the form

\begin{matrix} Δ_{g} = Δ_{F} + Δ_{H} + Δ_{- 1, 2} + K_{1} + K_{2} + K_{3}, \end{matrix}

(2.7)

where

$∙$ $Δ_{F} = d_{F} δ_{F} + δ_{F} d_{F} \in D^{0, 2} (M, ℱ, Λ T^{*} M)$ is the tangential Laplacian.
$∙$ $Δ_{H} = d_{H} δ_{H} + δ_{H} d_{H} \in D^{2, 0} (M, ℱ, Λ T^{*} M)$ is the transversal Laplacian.
$∙$ $Δ_{- 1, 2} = θ θ^{*} + θ^{*} θ \in D^{0, 0} (M, ℱ, Λ T^{*} M)$ .
$∙$ $K_{1} = d_{F} δ_{H} + δ_{H} d_{F} + δ_{F} d_{H} + d_{H} δ_{F} \in D^{1, 0} (M, ℱ, Λ T^{*} M)$ .
$∙$ $K_{2} = d_{F} θ^{*} + θ^{*} d_{F} + δ_{F} θ + θ δ_{F} \in D^{0, 0} (M, ℱ, Λ T^{*} M) .$
$∙$ $K_{3} = d_{H} θ^{*} + θ^{*} d_{H} + δ_{H} θ + θ δ_{H} \in D^{1, 0} (M, ℱ, Λ T^{*} M)$ .

We also introduce the first order differential operator

D_{H} = d_{H} + d_{H}^{*}

C^{\infty} (M, \land H^{*})

, which is called the transverse signature operator.

A basic property of geometric operators on manifolds equipped with Riemannian foliation is that, if

ℱ

is a Riemannian foliation and

g_{M}

is a bundle-like metric, then the operators

d_{F} δ_{H} + δ_{H} d_{F}

and

δ_{F} d_{H} + d_{H} δ_{F}

belong to

D^{0, 1} (M, ℱ, Λ T^{*} M)

. In particular,

K_{1} \in D^{0, 1} (M, ℱ, Λ T^{*} M)

3 Operator algebras of foliations

In this Section, we will describe the noncommutative algebras associated with the leaf space of a foliation. First, we will define an algebra, consisting of very nice functions, on which all basic operators of analysis are defined, then, depending on a problem in question, we will complete this algebra and obtain an analogue of the algebra of measurable, continuous or smooth functions. The role of a “nice” algebra is played by the algebra

C_{c}^{\infty} (G)

of smooth compactly supported functions on the holonomy groupoid

G

of the foliation. Therefore, we start with the notion of holonomy groupoid of a foliation.

3.1 Holonomy groupoid

A foliation

ℱ

defines an equivalence relation

ℛ \subset M \times M

M

(x, y) \in ℛ

if and only if

x

and

y

lie on the same leaf of the foliation

ℱ

. Generally,

ℛ

is not a smooth manifold, but one can resolve its singularity, constructing a smooth manifold

G

, called the holonomy groupoid or the graph of the foliation, which “almost every where” coincides with

ℛ

and which can be used in many cases as a substitution for

ℛ

. The idea of the holonomy groupoid appeared in the papers of Ehresmann, Reeb and Thom (cf. [67, 176] ) and was completely realized by Winkelnkemper [184] . First of all, we give the general definition of a groupoid (see [156, 137, 31, 126, 146] for groupoids and related subjects).

Definition 3.1. We say that a set

G

has the structure of a groupoid with the set of units

G^{(0)}

, if there are defined maps

$∙$ $Δ : G^{(0)} \to G$ (the diagonal map or the unit map);
$∙$ an involution $i : G \to G$ called the inversion and written as $i (γ) = γ^{- 1}$ ;
$∙$ a range map $r : G \to G^{(0)}$ and a source map $s : G \to G^{(0)}$ ;
$∙$ an associative multiplication $m : (γ, γ^{'}) \to γ γ^{'}$ defined on the set $G^{(2)} = {(γ, γ^{'}) \in G \times G : r (γ^{'}) = s (γ)},$

satisfying the conditions

$∙$ $r (Δ (x)) = s (Δ (x)) = x$ and $γ Δ (s (γ)) = γ$ , $Δ (r (γ)) γ = γ$ ;
$∙$ $r (γ^{- 1}) = s (γ)$ and $γ γ^{- 1} = Δ (r (γ))$ .

Alternatively, one can define a groupoid as a small category, where each morphism is an isomorphism.

It is convenient to think of an element

γ \in G

as an arrow

γ : x \to y

, going from

x = s (γ)

y = r (γ)

We will use the standard notation (for

x, y \in G^{(0)}

$∙$ $G^{x} = {γ \in G : r (γ) = x} = r^{- 1} (x)$ ,
$∙$ $G_{x} = {γ \in G : s (γ) = x} = s^{- 1} (x)$ ,
$∙$ $G_{y}^{x} = {γ \in G : s (γ) = x, r (γ) = y}$ .

Definition 3.2. A groupoid

G

is called smooth (or a Lie groupoid), if

G^{(0)}

G

and

G^{(2)}

are smooth manifolds,

r

s

i

and

m

are smooth maps,

r

and

s

are submersions, and

Δ

is an immersion.

Example 3.3. Lie groups. A Lie group

H

defines a smooth groupoid as follows:

G = H

G^{(0)}

consists of a single point, the maps

i

and

m

are given by the group operations in

H

Example 3.4. Trivial groupoid. Let

X

be an arbitrary set. Put

G = X

G^{(0)} = X

, the maps

s

and

r

are the identity maps (that is, in other words, each element

x \in G^{(0)} = X

is identified with an unique element

γ : x \to x

Example 3.5. Equivalence relations. Any equivalence relation

R \subset X \times X

defines a groupoid, if we put

G^{(0)} = X

G = R

, the maps

s : R \to X

and

r : R \to X

are given by

s (x, y) = y

r (x, y) = x

. Thus, pairs

(x_{1}, y_{1})

and

(x_{2}, y_{2})

can be multiplied if and only if

y_{1} = x_{2}

, and in this case

(x_{1}, y_{1}) (x_{2}, y_{2}) = (x_{1}, y_{2})

In the particular case of

R = X \times X

, we get a so called principal or pair groupoid.

Example 3.6. Group actions. Let a Lie group

H

act smoothly from the left on a smooth manifold

X

. The crossed product groupoid

X ∖ H

is defined as follows:

G^{(0)} = X

G = X \times H

. The maps

s : X \times H \to X

and

r : X \times H \to X

have the form

s (x, h) = h^{- 1} x

r (x, h) = x

. Thus, pairs

(x_{1}, h_{1})

and

(x_{2}, h_{2})

can be multiplied if and only if

x_{2} = h_{1}^{- 1} x_{1}

, and in this case

(x_{1}, h_{1}) (x_{2}, h_{2}) = (x_{1}, h_{1} h_{2})

Example 3.7. Fundamental groupoid (cf. for instance, [169] ). Let

X

be a topological space,

G = Π (X)

the set of homotopy classes of paths in

X

with all possible endpoints. More precisely, if

γ : [0, 1] \to X

is a path from

x = γ (0)

y = γ (1)

, then we denote by

[γ]

the homotopy class of

γ

with fixed

x

and

y

. Define the groupoid

Π (X)

as the set of triples

(x, [γ], y)

, where

x, y \in X

γ

is a path with the initial point

x = γ (0)

and the final point

y = γ (1)

, where the multiplication is given by the product of paths. The groupoid

Π (X)

is called the fundamental groupoid of

X

Example 3.8. The Haefliger groupoid

Γ_{n}

[23, 89, 91] . Let

M

be a smooth manifold. A groupoid

Γ_{M}

consists of the germs of local diffeomorphisms of

M

at arbitrary points of

M

(Γ_{M})^{(0)} = M

. If

γ \in Γ_{M}

is the germ at

x \in M

of a diffeomorphism

f

from some neighborhood

U

x

on an open set

f (U)

, then

s (γ) = x

r (γ) = f (x)

The multiplication in

Γ_{M}

is given by the composition of maps. If

M = R^{n}

, then the groupoid

Γ_{M}

is denoted by

Γ_{n}

The holonomy groupoid

G = G (M, ℱ)

of a foliated manifold

(M, ℱ)

is defined in the following way. Let

\sim_{h}

be an equivalence relation on the set of continuous leafwise paths

γ : [0, 1] \to M

, setting

γ_{1} \sim_{h} γ_{2}

, if

γ_{1}

and

γ_{2}

have the same initial and final points and the same holonomy maps:

h_{γ_{1}} = h_{γ_{2}}

. The holonomy groupoid

G

is the set of

\sim_{h}

-equivalence classes of leafwise paths. The set of units

G^{(0)}

is a manifold

M

. The multiplication in

G

is given by the product of paths. The corresponding range and source maps

s, r : G \to M

are given by

s (γ) = γ (0)

and

r (γ) = γ (1)

. Finally, the diagonal map

Δ : M \to G

takes any

x \in M

to the element in

G

given by the constant path

γ (t) = x, t \in [0, 1]

. To simplify the notation, we will identify

x \in M

with

Δ (x) \in G

For any

x \in M

the map

s

maps

G^{x}

on the leaf

L_{x}

through

x

. The group

G_{x}^{x}

coincides with the holonomy group of

L_{x}

. The map

s : G^{x} \to L_{x}

is the covering map associated with the group

G_{x}^{x}

, called the holonomy covering. One can also introduce the holonomy groupoid

G (L)

of a leaf

L

ℱ

as the set of

\sim_{h}

-equivalence classes of piecewise smooth paths in

L

The holonomy groupoid

G

has the structure of a smooth (in general, non-Hausdorff and non-paracompact) manifold of dimension

2 p + q

. Recall the construction of an atlas on

G

[34] .

Let

φ : U \to I^{p} \times I^{q}, φ^{'} : U^{'} \to I^{p} \times I^{q}

be two foliated charts,

π = p r_{n q} \circ φ : U \to R^{q}

π^{'} = p r_{n q} \circ φ^{'} : U^{'} \to R^{q}

the corresponding distinguished maps. The foliated charts

φ

φ^{'}

are called compatible, if, for any

m \in U

and

m^{'} \in U^{'}

with

π (m) = π^{'} (m^{'})

, there is a leafwise path

γ

from

m

m^{'}

such that the corresponding holonomy map

h_{γ}

takes the germ

π_{m}

π

m

to the germ

π_{m^{'}}^{'}

π^{'}

m^{'}

For any pair of compatible foliated charts

φ

and

φ^{'}

denote by

W (φ, φ^{'})

the subset in

G

, consisting of all

γ \in G

from

s (γ) = m = φ^{- 1} (x, y) \in U

r (γ) = m^{'} = {φ^{'}}^{- 1} (x^{'}, y) \in U^{'}

such that the corresponding holonomy map

h_{γ}

takes the germ

π_{m}

of the map

π = p r_{n q} \circ φ

m

to the germ

π_{m^{'}}^{'}

of the map

π^{'} = p r_{n q} \circ φ^{'}

m^{'}

. There is a coordinate map

\begin{matrix} Γ : W (φ, φ^{'}) \to I^{p} \times I^{p} \times I^{q}, \end{matrix}

(3.1)

which takes each element

γ \in W (φ, φ^{'})

such that

s (γ) = m = ϰ^{- 1} (x, y)

r (γ) = m^{'} = {φ^{'}}^{- 1} (x^{'}, y)

and

h_{γ} π_{m} = π_{m^{'}}^{'}

to the triple

(x, x^{'}, y) \in I^{p} \times I^{p} \times I^{q}

As shown in [34] , the coordinate neighborhoods

W (φ, φ^{'})

form an atlas of a

(2 p + q)

-dimensional manifold (in general, non-Hausdorff and non-paracompact) on

G

. Moreover, the groupoid

G

is a smooth groupoid.

Non-Hausdorffness of the holonomy groupoid is related with the phenomenon of one-sided holonomy. The simplest example of a foliation with the non-Hausdorff holonomy groupoid is given by the trajectories of a nonsingular vector field on the plane, having a one-sided limit cycle. As shown in [184] , the holonomy groupoid is Hausdorff if and only if the holonomy maps

H_{T_{0} T_{1}} (γ_{1})

and

H_{T_{0} T_{1}} (γ_{2})

along any leafwise paths

γ_{1}

and

γ_{2}

with the initial point

x

and the final point

y

, given by smooth transversals

T_{0}

and

T_{1}

, passing through

x

and

y

accordingly, coincide, if they coincide on some open subset

U \subset T_{0}

such that

x \in \bar{U}

. In particular, the holonomy groupoid is Hausdorff, if the holonomy is trivial, or real analytic. Moreover, the holonomy groupoid of a Riemannian foliation is Hausdorff. In the sequel, we will always assume that

G

is a Hausdorff manifold.

Example 3.9. If a simple foliation

ℱ

is defined by a submersion

π : M \to B

, then its holonomy groupoid

G

consists of all

(x, y) \in M \times M

such that

π (x) = π (y)

, and, moreover,

G^{(0)} = M

, the maps

s : G \to M

and

r : G \to M

are given by

s (x, y) = y

r (x, y) = x

Example 3.10. If a foliation

ℱ

is given by the orbits of a free smooth action of a connected Lie group

H

on a manifold

M

, then its holonomy groupoid coincides with the crossed product groupoid

M ∖ H

Besides the holonomy groupoid, there are another groupoids, which can be associated with the foliation. First of all, it is the groupoid given by the equivalence relation on

M

, setting points

x

and

y

to be equivalent, if they lie on the same leaf of the foliation (the coarse groupoid). As noted above, this groupoid is not smooth. One can also consider the fundamental groupoid of the foliation

Π (M, ℱ)

, which also consists of equivalence classes of leafwise paths, where two leafwise paths are called equivalent, if they are homotopic in the class of leafwise paths with fixed endpoints. The fundamental groupoid of the foliation

Π (M, ℱ)

is a smooth groupoid (cf., for instance, [149] ).

There is a foliation

G

of dimension

2 p

on the holonomy groupoid

G

. In any coordinate chart

W (φ, φ^{'})

given by a pair of compatible foliated charts

φ

and

φ^{'}

, the leaves of

G

are given by equations of the form

y = c o n s t

. The leaf of

G

through

γ \in G

consists of all

γ^{'} \in G

such that

r (γ)

and

r (γ^{'})

lie on the same leaf of

ℱ

and coincides with the holonomy groupoid of this leaf.

The holonomy group of a leaf of

G

coincides with the holonomy group of the corresponding leaf of

ℱ

. (The last statement corrects an erroneous one made in [184] . This fact was noted, for instance, by Molino in his review of [184] in Mathematical Reviews (see MR 85j:57043).) The differential of the map

(r, s) : G \to M \times M

maps isomorphically the tangent bundle

T G

G

to the bundle

F ⊠ F

M \times M

, therefore, there is a canonical isomorphism

T G \sim = r^{*} F \oplus s^{*} F

A distribution

H

M

transverse to

ℱ

determines a distribution

H G

G

transverse to

G

. For any

X \in H_{y}

, there is a unique vector

\hat{X} \in T_{γ} G

such that

d s (\hat{X}) = d h_{γ}^{- 1} (X)

and

d r (\hat{X}) = X

, where

d h_{γ} : H_{x} \to H_{y}

is the linear holonomy map associated with

γ

. The space

H_{γ} G

consists of all vectors of the form

\hat{X} \in T_{γ} G

for different

X \in H_{y}

. In any coordinate chart

W (φ, φ^{'})

G

, the tangent space

T_{γ} G

G

at some

γ

with the coordinates

(x, x^{'}, y)

consists of vectors of the form

X \frac{\partial}{\partial x} + X^{'} \frac{\partial}{\partial x^{'}}

and the distribution

H_{γ} G

consists of vectors

X \frac{\partial}{\partial x} + X^{'} \frac{\partial}{\partial x^{'}} + Y \frac{\partial}{\partial y}

such that

X \frac{\partial}{\partial x} + Y \frac{\partial}{\partial y} \in H_{(x, y)}

and

X^{'} \frac{\partial}{\partial x^{'}} + Y \frac{\partial}{\partial y} \in H_{(x^{'}, y)}

Let

g_{M}

be a Riemannian metric on

M

and

H = F^{⊥}

. Then a Riemannian metric

g_{G}

G

is defined as follows. All the components in

T_{γ} G = F_{y} \oplus F_{x} \oplus H_{γ} G

are mutually orthogonal, and, by definition,

g_{G}

coincides with

g_{M}

F_{y} \oplus H_{γ} G \sim = F_{y} \oplus H_{y} = T_{y} M

and with

g_{F}

F_{x}

ℱ

is Riemannian and

g_{M}

is a bundle-like Riemannian metric, then

g_{G}

is bundle-like, and, therefore,

G

is Riemannian. Moreover, in this case the maps

s : G \to M

and

r : G \to M

are Riemannian submersions and locally trivial fibrations. In particular, the holonomy coverings

G^{x}

of leaves of

ℱ

are diffeomorphic [184] .

3.2 The $C^{*}$ -algebra of a foliation and noncommutative topology

In this Section, we will describe the construction of the

C^{*}

-algebra associated with an arbitrary foliation, which is an analogue of the algebra of continuous functions on the leaf space of the foliation.

Definition 3.11. (cf., for instance, [62, 141, 147, 174] ) A

C^{*}

-algebra is an involutive Banach algebra

A

such that

∥ a^{*} a ∥ = ∥ a ∥^{2}, a \in A .

Example 3.12. The simplest example of a

C^{*}

-algebra is given by the algebra

C_{0} (X)

of continuous functions on a locally compact Hausdorff topological space

X

, vanishing at the infinity, which is endowed with operations of the pointwise addition and the multiplication, with the standard involution and with the uniform norm

∥ f ∥ = {sup}_{x \in X} | f (x) |, f \in C_{0} (X) .

The Gelfand-Naimark theorem allows to reconstruct uniquely from a commutative

C^{*}

-algebra

A

the locally compact Hausdorff topological space

X

such that

A \sim = C_{0} (X)

. More precisely,

X

coincides with the set

\hat{A}

of all characters of the algebra

A

, i. e., of all continuous homomorphisms

A \to C

, endowed with the topology of pointwise convergence.

The previous example permits to consider an arbitrary

C^{*}

-algebra as the algebra of continuous functions on some virtual space. By this reason, the theory of

C^{*}

-algebras is often called as noncommutative topology.

Example 3.13. The algebra

ℒ (H)

of bounded operators in a Hilbert space

H

equipped with the involution given by taking the adjoints and with the operator norm is a

C^{*}

-algebra.

By the second Gelfand-Naimark theorem, any

C^{*}

-algebra is isometrically

*

-isomorphic to some norm closed

*

-subalgebra of the algebra

ℒ (H)

for some Hilbert space

H

There are two ways to define the

C^{*}

-algebras associated with a foliation.

The first way makes use of the auxiliary choices of a smooth Haar system, the second one requires no auxiliary choices and uses the language of half-densities.

3.2.1 Definitions, using a Haar system.

In this Section we give the definition of the

C^{*}

-algebras associated with an arbitrary smooth groupoid

G

. In fact, the assumption of smoothness of a groupoid is not essential here, and all the definitions can be generalized to the case of topological groupoids (cf., for instance, [156] ).

We will only consider Hausdorff groupoids. For the definition of the

C^{*}

-algebra of a foliation in the case when the holonomy groupoid is non-Hausdorff, cf. [37] (and also [57, 58, 60] ).

Definition 3.14. A smooth Haar system on a smooth groupoid

G

is a family of positive Radon measures

{ν^{x} : x \in G^{(0)}}

G

, satisfying the following conditions:

(1) The support of the measure $ν^{x}$ coincides with $G^{x}$ , and $ν^{x}$ is a smooth measure on $G^{x}$ .
(2) The family ${ν^{x} : x \in G^{(0)}}$ is left-invariant, that is, for any continuous function $f \in C (G^{x}), f \geq 0,$ and for any $γ \in G, s (γ) = x, r (γ) = y,$ we have $\int_{G^{y}} f (γ_{1}) d ν^{y} (γ_{1}) = \int_{G^{x}} f (γ γ_{1}) d ν^{x} (γ_{1}) .$
(3) The family ${ν^{x} : x \in G^{(0)}}$ is smooth, that is, for any $φ \in C_{c}^{\infty} (G)$ the function $x \in G^{(0)} \mapsto \int_{G^{x}} φ (γ) d ν^{x} (γ)$ is a smooth function on $G^{(0)}$ .

For a compact foliated manifold

(M, ℱ)

, there is a distinguished class of smooth Haar systems

{ν^{x} : x \in G^{(0)}}

on the holonomy groupoid

G

ℱ

given by smooth positive leafwise densities

α \in C^{\infty} (M, | T ℱ |)

. The positive Radon measure

ν^{x}

G^{x}, x \in M,

is given as the lift of the density

α

via the holonomy covering

s : G^{x} \to M

. In the sequel, we will assume that a Haar system on

G

constructed in such a way is fixed.

Let

G

be a smooth groupoid,

G^{(0)} = M

and

{ν^{x} : x \in M}

a smooth Haar system. Introduce the structure of an involutive algebra on

C_{c}^{\infty} (G)

\begin{matrix} k_{1} * k_{2} (γ) & = \int_{G^{x}} k_{1} (γ_{1}) k_{2} (γ_{1}^{- 1} γ) d ν^{x} (γ_{1}), γ \in G^{x}, \end{matrix}

\begin{matrix} k^{*} (γ) & = \bar{k (γ^{- 1})}, γ \in G . \end{matrix}

\begin{matrix}  \end{matrix}

For any

x \in M

, there is a natural representation of

C_{c}^{\infty} (G)

in the Hilbert space

L^{2} (G^{x}, ν^{x})

given, for

k \in C_{c}^{\infty} (G)

and

ζ \in L^{2} (G^{x}, ν^{x})

, by

R_{x} (k) ζ (γ) = \int_{G^{x}} k (γ^{- 1} γ_{1}) ζ (γ_{1}) d ν^{x} (γ_{1}), r (γ) = x .

The completion of the involutive algebra

C_{c}^{\infty} (G)

in the norm

∥ k ∥ = {sup}_{x} ∥ R_{x} (k) ∥

is called the reduced

C^{*}

-algebra of the groupoid

G

and denoted by

C_{r}^{*} (G)

There is also defined the full

C^{*}

-algebra of the groupoid

C^{*} (G)

, which is the completion of

C_{c}^{\infty} (G)

in the norm

∥ k ∥_{max} = sup ∥ π (k) ∥,

where supremum is taken over the set of all

*

-representations

π

of the algebra

C_{c}^{\infty} (G)

in Hilbert spaces.

Any

k \in C_{c}^{\infty} (G)

defines a bounded operator

R (k)

C_{c}^{\infty} (M)

. For any

u \in C_{c}^{\infty} (M)

, we have

R (k) u (x) = \int_{G^{x}} k (γ) u (s (γ)) d ν^{x} (γ), x \in M .

The correspondence

k \mapsto R (k)

defines a representation of the algebra

C_{c}^{\infty} (G)

C_{c}^{\infty} (M)

Example 3.15. Lie groups. If a groupoid

G

is given by a Lie group

H

, then a smooth Haar system on

G

is given by a left-invariant Haar measure

d h

H

, and the multiplication in

C_{c}^{\infty} (G)

is the convolution operation defined for any

u, v \in C_{c}^{\infty} (H)

(u * v) (g) = \int_{H} u (h) v (h^{- 1} g) d h, g \in H,

the involution is given by

u^{*} (g) = \bar{u (g^{- 1})}, u \in C_{c}^{\infty} (H),

and the operator algebras

C_{r}^{*} (G)

and

C^{*} (G)

are the group

C^{*}

-algebras

C_{r}^{*} (H)

and

C^{*} (H)

(cf., for instance, [147] ).

Example 3.16. Trivial groupoid. Let

X

be a smooth manifold.

Put

G = X

G^{(0)} = X

, the maps

s

and

r

are the identity maps. The operator algebras

C_{r}^{*} (G)

and

C^{*} (G)

coincide with the commutative

C^{*}

-algebra

C_{0} (X)

Example 3.17. Principal groupoid. Let

X

be a smooth manifold.

Put

G^{(0)} = X

G = X \times X

. The maps

s : X \times X \to X

and

r : X \times X \to X

are given by

s (x, y) = y

r (x, y) = x

. If we choose a Haar system on

G

, taking as

ν^{x}

a fixed smooth measure

μ

in each

G^{x} = X

, then the operations in

C_{c}^{\infty} (G)

are given by

\begin{matrix} (k_{1} * k_{2}) (x, y) & = \int_{X} k_{1} (x, z) k_{2} (z, y) d μ (z), (x, y) \in X \times X, \end{matrix}

\begin{matrix} k^{*} (x, y) & = \bar{k (y, x)}, (x, y) \in X \times X, \end{matrix}

\begin{matrix}  \end{matrix}

where

k, k_{1}, k_{2} \in C_{c}^{\infty} (G)

. Thus, elements of

C_{c}^{\infty} (G)

can be considered as the kernels of integral operators in

C^{\infty} (X)

. The representation

k \to R_{x} (k)

takes each

k \in C_{c}^{\infty} (G) \subset C^{\infty} (X \times X)

to the integral operator in

L^{2} (G^{x}, ν^{x}) \sim = L^{2} (X, μ)

with the integral kernel

k

R_{x} (k) u (y) = \int_{X} k (y, z) u (z) d μ (z), u \in L^{2} (X, μ) .

Finally, the operator algebras

C_{r}^{*} (G)

and

C^{*} (G)

coincide with the algebra

C_{0} (X) \otimes K (L^{2} (X, μ))

Example 3.18. Group actions. Let a Lie group

H

act smoothly from the left on a smooth manifold

X

. Consider the corresponding crossed product groupoid

G = X ∖ H

. Then

G^{x} = H

for any

x \in X

and a smooth Haar system on

G = X ∖ H

is given by a left-invariant Haar measure

d h

H

. The multiplication in

C_{c}^{\infty} (G)

is defined for any

u, v \in C_{c}^{\infty} (X \times H)

(u * v) (x, g) = \int_{H} u (x, h) v (h^{- 1} x, h^{- 1} g) d h, (x, g) \in X \times H,

the involution is defined for

u \in C_{c}^{\infty} (X \times H)

u^{*} (x, g) = \bar{u (g^{- 1} x, g^{- 1})}, (x, g) \in X \times H .

The operator algebras

C_{r}^{*} (G)

and

C^{*} (G)

associated with the crossed product groupoid

G = X ∖ H

coincide with the crossed products

C_{0} (X) ∖_{r} H

and

C_{0} (X) ∖ H

of the algebra

C_{0} (X)

by the group

H

with respect to the induced action of

H

C_{0} (X)

(cf., for instance, [147] ).

If the group

H

is discrete, elements of

C_{c}^{\infty} (G)

are families

{a_{γ} \in C^{\infty} (X) : γ \in H}

such that

a_{γ} \neq 0

for finitely many elements

γ

. It is convenient to write them as

a = \sum_{γ \in H} a_{γ} U_{γ}

. The multiplication in

C_{c}^{\infty} (G)

is written as

(a_{γ_{1}} U_{γ_{1}}) (b_{γ_{2}} U_{γ_{2}}) = (a_{γ_{1}} T_{γ_{1}} (b_{γ_{2}})) U_{γ_{1} γ_{2}},

where

T_{γ}

denotes the operator in

C_{0} (X)

induced by the action of

γ \in H

T_{γ} f (x) = f (γ^{- 1} x), x \in X, f \in C_{0} (X) .

The involution in the algebra

C_{c}^{\infty} (G)

is given by

(a_{γ} U_{γ})^{*} = {\bar{a}}_{γ} U_{γ^{- 1}} .

Let

G

be the holonomy groupoid of a foliation

ℱ

on a compact manifold

M

. Elements of the algebra

C_{c}^{\infty} (G)

can be considered as families of the kernels of integral operators along the leaves of the foliation (more precisely, on the holonomy coverings

G^{x}

). Namely, each

k \in C_{c}^{\infty} (G)

corresponds to the family

{R_{x} (k) : x \in M}

, where

R_{x} (k)

is the integral operator in

L^{2} (G^{x}, ν^{x})

given by the integral kernel

K (γ_{1}, γ_{2}) = k (γ_{1}^{- 1} γ_{2}), γ_{1}, γ_{2} \in G^{x} .

The product of elements

k_{1}

and

k_{2}

from

C_{c}^{\infty} (G)

corresponds to the composition of integral operators

{R_{x} (k_{1}) R_{x} (k_{2}) : x \in M}

. Finally, the representation

R

means the natural action of such families of leafwise integral operators in

L^{2} (M)

. In this case, the algebra

C_{r}^{*} (G)

will be often called the reduced

C^{*}

-algebra of the foliation and denoted by

C^{*} (M, ℱ)

Example 3.19. Consider the simplest example of a foliation, given by the linear foliation on the torus. Thus, suppose that

M = T^{2} = R^{2} / Z^{2}

is the two-dimensional torus and a foliation

ℱ_{θ}

is given by the trajectories of the vector field

X = \frac{\partial}{\partial x} + θ \frac{\partial}{\partial y},

where

θ \in R

is a fixed irrational number.

Since this foliation is given by the orbits of a free group action of

R

T^{2}

, its holonomy groupoid coincides with the crossed product groupoid

T^{2} ∖ R

. Thus,

G = T^{2} \times R

G^{(0)} = T^{2}

s (x, y, t) = (x - t, y - θ t)

r (x, y, t) = (x, y)

(x, y) \in T^{2}, t \in R,

and the multiplication is given by

(x_{1}, y_{1}, t_{1}) (x_{2}, y_{2}, t_{2}) = (x_{1}, y_{1}, t_{1} + t_{2}),

x_{2} = x_{1} - t_{1}, y_{2} = y_{1} - θ t_{1}

The

C^{*}

-algebra

C_{r}^{*} (G)

of the linear foliation on

T^{2}

coincides with the crossed product

C (T^{2}) ∖_{r} R

. Therefore, the product

k_{1} * k_{2}

k_{1}, k_{2} \in C_{c}^{\infty} (T^{2} \times R) \subset C (T^{2}) ∖_{r} R

is given by

(k_{1} * k_{2}) (x, y, t) = \int_{- \infty}^{\infty} k_{1} (x_{1}, y_{1}, t_{1}) k_{2} (x - t_{1}, y - θ t_{1}, t - t_{1}) d t_{1}, (x, y) \in T^{2}, t \in R,

and, for any

k \in C_{c}^{\infty} (T^{2} \times R)

k^{*} (x, y, t) = \bar{k (x - t, y - θ t, - t)}, (x, y) \in T^{2}, t \in R .

For any

k \in C_{c}^{\infty} (T^{2} \times R)

and for any

(x, y) \in T^{2}

, the operator

R_{(x, y)} (k)

L^{2} (G^{(x, y)}) \sim = L^{2} (R)

has the form: for any

u \in L^{2} (R)

R_{(x, y)} (k) u (t) = \int_{- \infty}^{\infty} k (x - t_{1}, y - θ t_{1}, t - t_{1}) u (t_{1}) d t_{1}, t \in R .

Finally, for any

k \in C_{c}^{\infty} (T^{2} \times R)

, the corresponding operator

R (k)

L^{2} (T^{2})

is given by

R (k) u (x, y) = \int_{- \infty}^{\infty} k (x, y, t) u (x - t, y - θ t) d t, (x, y) \in T^{2} .

θ

is rational, then the linear foliation on

T^{2}

is given by the orbits of a free group action of

S^{1}

T^{2}

, and its holonomy groupoid coincides with the crossed product groupoid

G = T^{2} ∖ S^{1}

We will also need the twisted version of the notion of the

C^{*}

-algebra of a foliation

(M, ℱ)

, with coefficients in a Hermitian vector bundle

E

M

. Denote by

C_{c}^{\infty} (G, ℒ (E))

the space of smooth, compactly supported sections of the vector bundle

(s^{*} E)^{*} \otimes r^{*} E

G

. In other words, the value of

k \in C_{c}^{\infty} (G, ℒ (E))

at any

γ \in G

is a linear map

k (γ) : E_{s (γ)} \to E_{r (γ)} .

The structure of an involutive algebra on

C_{c}^{\infty} (G, ℒ (E))

is defined by analogous formulas

\begin{matrix} k_{1} * k_{2} (γ) & = \int_{G^{x}} k_{1} (γ_{1}) k_{2} (γ_{1}^{- 1} γ) d ν^{x} (γ_{1}), γ \in G, r (γ) = x, \end{matrix}

\begin{matrix} k^{*} (γ) & = k (γ^{- 1})^{*}, γ \in G . \end{matrix}

\begin{matrix}  \end{matrix}

with the only difference that the product

k_{1} (γ_{1}) k_{2} (γ_{1}^{- 1} γ)

means here the composition of the linear maps

k_{2} (γ_{1}^{- 1} γ) : E_{s (γ)} \to E_{s (γ_{1})}

and

k_{1} (γ_{1}) : E_{s (γ_{1})} \to E_{r (γ_{1})}

and

k (γ^{- 1})^{*}

means the adjoint of

k (γ^{- 1}) : E_{r (γ)} \to E_{s (γ)} .

Consider the vector bundle

\tilde{E} = s^{*} (E)

G

. Let

{\tilde{E}}^{x}

be the restriction of the bundle

\tilde{E}

G^{x}

, and

L^{2} (G^{x}, {\tilde{E}}^{x})

the Hilbert space of

L^{2}

-sections of the bundle

{\tilde{E}}^{x}

, determined by the fixed Hermitian structure on

{\tilde{E}}^{x}

and the measure

ν^{x}

. For any

x \in M

, we introduce a representation

R_{x}

of the algebra

C_{c}^{\infty} (G, ℒ (E))

L^{2} (G^{x}, {\tilde{E}}^{x})

, given, for

k \in C_{c}^{\infty} (G, ℒ (E))

, by

\begin{matrix} R_{x} (k) ζ (γ) = \int_{G^{x}} k (γ^{- 1} γ_{1}) ζ (γ_{1}) d ν^{x} (γ_{1}), ζ \in L^{2} (G^{x}, {\tilde{E}}^{x}) . \end{matrix}

(3.2)

The completion of

C_{c}^{\infty} (G, ℒ (E))

in the norm

∥ k ∥ = {sup}_{x} ∥ R_{x} (k) ∥

is called the reduced (twisted)

C^{*}

-algebra of the foliation with coefficients in

E

and denoted by

C_{r}^{*} (G, E)

C^{*} (M, ℱ, E)

. Denote by

C^{*} (G, E)

the full

C^{*}

-algebra of the foliation, defined in the same way as in the case of trivial coefficients.

Any

k \in C_{c}^{\infty} (G, ℒ (E))

defines an operator

R_{E} (k)

C^{\infty} (M, E)

. For any

u \in C^{\infty} (M, E)

, we have

\begin{matrix} R_{E} (k) u (x) = \int_{G^{x}} k (γ) u (s (γ)) d ν^{x} (γ), x \in M . \end{matrix}

(3.3)

The correspondence

k \mapsto R_{E} (k)

determines a representation of the algebra

C_{c}^{\infty} (G, ℒ (E))

L^{2} (M, E)

Let us give some facts, which relate the structure of the reduced

C^{*}

-algebra of a foliation

C^{*} (M, ℱ)

with the topology of

ℱ

(for more details cf. [68, 99] ).

Theorem 3.20 ([68] ). Let

(M, ℱ)

be a foliated manifold.

(1) The $C^{*}$ -algebra $C^{*} (M, ℱ)$ is simple if and only if $ℱ$ is minimal, i.e. every its leaf is dense in $M$ .
(2) The $C^{*}$ -algebra $C^{*} (M, ℱ)$ is primitive if and only if $ℱ$ is (topologically) transitive, i.e. it has a leaf, dense in $M$ .
(3) If $ℱ$ is amenable in the sense that $C^{*} (M, ℱ) = C^{*} (G)$ , the $C^{*}$ -algebra $C^{*} (M, ℱ)$ has a representation, consisting of compact operators if and only if $ℱ$ has a compact leaf.

In the work [68] , a description of the space of primitive ideals of the

C^{*}

-algebra

C^{*} (M, ℱ)

is also given.

3.2.2 Definition, using half-densities

In this Section, we will give the definitions of the operator algebras associated with a foliated manifold, which make no choice of a Haar system.

For this, we will use the language of half-densities.

Let

(M, ℱ)

be a compact foliated manifold. Consider the vector bundle of leafwise half-densities

| T ℱ |^{1 / 2}

M

. Pull back

| T ℱ |^{1 / 2}

to the vector bundles

s^{*} (| T ℱ |^{1 / 2})

and

r^{*} (| T ℱ |^{1 / 2})

on the holonomy groupoid

G

, using the source map

s

and the range map

r

. Define a vector bundle

| T G |^{1 / 2}

G

| T G |^{1 / 2} = r^{*} (| T ℱ |^{1 / 2}) \otimes s^{*} (| T ℱ |^{1 / 2}) .

The bundle

| T G |^{1 / 2}

is naturally identified with the bundle of leafwise half-densities on the foliated manifold

(G, G)

The structure of an involutive algebra on

C_{c}^{\infty} (G, | T G |^{1 / 2})

is defined as

\begin{matrix} σ_{1} * σ_{2} (γ) & = \int_{γ_{1} γ_{2} = γ} σ_{1} (γ_{1}) σ_{2} (γ_{2}), γ \in G, \end{matrix}

\begin{matrix} σ^{*} (γ) & = \bar{σ (γ^{- 1})}, γ \in G, \end{matrix}

\begin{matrix}  \end{matrix}

where

σ, σ_{1}, σ_{2} \in C_{c}^{\infty} (G, | T G |^{1 / 2})

. The formula for

σ_{1} * σ_{2}

should be interpreted in the following way. If we write

γ : x \to y, γ_{1} : z \to y

and

γ_{2} : x \to z

, then

\begin{matrix} σ_{1} (γ_{1}) σ_{2} (γ_{2}) \in & | T_{y} ℱ |^{1 / 2} \otimes | T_{z} ℱ |^{1 / 2} \otimes | T_{z} ℱ |^{1 / 2} \otimes | T_{x} ℱ |^{1 / 2} \end{matrix}

\begin{matrix} \sim = | T_{y} ℱ |^{1 / 2} \otimes | T_{z} ℱ |^{1} \otimes | T_{x} ℱ |^{1 / 2}, \end{matrix}

\begin{matrix}  \end{matrix}

and, integrating the

| T_{z} ℱ |^{1}

-component

σ_{1} (γ_{1}) σ_{2} (γ_{2})

with respect to

z \in M

, we get a well-defined section of the bundle

r^{*} (| T ℱ |^{1 / 2}) \otimes s^{*} (| T ℱ |^{1 / 2}) = | T G |^{1 / 2} .

In the case of nontrivial coefficients, taking values in an Hermitian vector bundle

E

M

, it is necessary to consider the space

C_{c}^{\infty} (G, ℒ (E) \otimes | T G |^{1 / 2})

, where the structure of an involutive algebra is defined by the same formulas.

The formula ( 3.3 ) can be rewritten in the language of half-densities as follows. For any

σ \in C_{c}^{\infty} (G, ℒ (E) \otimes | T G |^{1 / 2})

and

u \in C^{\infty} (M, E \otimes | T ℱ |^{1 / 2})

, the element

R_{E} (σ) u

C^{\infty} (M, E \otimes | T ℱ |^{1 / 2})

is given by

R_{E} (σ) u (x) = \int_{G^{x}} σ (γ) s^{*} u (γ), x \in M .

This formula should be interpreted as follows. We have

s^{*} u \in C_{c}^{\infty} (G, s^{*} (E \otimes | T ℱ |^{1 / 2}))

and, hence,

σ \cdot s^{*} u \in C_{c}^{\infty} (G, r^{*} E \otimes r^{*} (| T ℱ |^{1 / 2}) \otimes s^{*} (| T ℱ |))

. The integration of the component in

s^{*} (| T ℱ |)

over

G^{x}

, i.e. with a fixed

r (γ) = x \in M

, gives a well-defined section

R_{E} (σ) u

E \otimes | T ℱ |^{1 / 2}

M

3.3 Von Neumann algebras and noncommutative measure theory

The initial datum of noncommutative measure theory is a pair

(ℳ, φ)

, consisting of a von Neumann algebra

ℳ

and a weight

φ

ℳ

Definition 3.21. A von Neumann algebra is an involutive subalgebra of the algebra

ℒ (H)

of bounded operators in a Hilbert space

H

, closed in the weak operator topology.

Definition 3.22. A weight on a von Neumann algebra

ℳ

is a function

φ

, defined on the set

ℳ_{+}

of positive elements of

ℳ

, with values in

{\bar{R}}_{+} = [0 . + \infty]

, satisfying the conditions

φ (a + b) = φ (a) + φ (b), a, b \in ℳ_{+}, φ (α a) = α φ (a), α \in R_{+}, a \in ℳ_{+} .

A weight on a von Neumann algebra

ℳ

is called a trace, if

φ (a^{*} a) = φ (a a^{*}), a \in ℳ_{+} .

Definition 3.23. A weight

φ

on a von Neumann algebra

ℳ

is called

(1) faithful, if, for any $a \in ℳ_{+}$ , the identity $φ (a) = 0$ implies $a = 0$ ;
(2) normal, if, for any bounded increasing net ${a_{α}}$ of elements from $ℳ_{+}$ with the least upper bound $a$ , the following identity holds: $φ (a) = {sup}_{α} φ (a_{α}) .$
(3) semifinite, if the linear span of the set ${x \in ℳ_{+} : φ (x) < \infty}$ is $σ$ -weakly dense in $ℳ$ .

Every von Neumann algebra has a faithful, normal, semifinite weight.

Example 3.24. The trace functional

t r

on the von Neumann algebra

ℒ (ℋ)

of bounded linear operators in a Hilbert space

ℋ

is a faithful, normal, semifinite trace.

Example 3.25. If a measure space

X

is endowed with a

σ

-finite measure

μ

, then elements of

L^{\infty} (X, μ)

considered as multiplication operators in the Hilbert space

L^{2} (X, μ)

form a von Neumann algebra.

Moreover, the identity

φ (f) = \int_{X} f (x) d μ (x), f \in L^{\infty} (X, μ),

defines a faithful, normal, semifinite trace on

L^{\infty} (X, μ)

The foundations of the noncommutative integration theory for foliations were laid by Connes in [34] . Let

(M, ℱ)

be a compact foliated manifold. Let

ρ

be a strictly positive continuous transverse density,

α

a strictly positive smooth leafwise density, and

ν = s^{*} α

the corresponding smooth Haar system.

The measures

ρ

and

α

can be combined to construct a measure

μ

M

Finally, the measure

μ

and the Haar system

ν

define a measure

m

G

\int_{G} f (γ) d m (γ) = \int_{M} (\int_{G^{x}} f (γ) d ν^{x} (γ)) d μ (x), f \in C_{c} (G) .

In the previous Section we defined the representation

R_{x}

of the involutive algebra

C_{c}^{\infty} (G)

in the Hilbert space

L^{2} (G^{x}, ν^{x})

for any

x \in M

. Consider a representation

R

of the algebra

C_{c}^{\infty} (G)

L^{2} (G, m) = \int_{M}^{\oplus} L^{2} (G^{x}, ν^{x}) d μ (x)

defined as the direct integral of the representations

R_{x}

R = \int_{M}^{\oplus} R_{x} d μ (x) .

Definition 3.26. The von Neumann algebra

W^{*} (M, ℱ)

of the foliation

ℱ

is defined as the closure of the image of

C_{c}^{\infty} (G)

under the representation

R

in the weak operator topology of

ℒ (L^{2} (G, m))

Since the definition of the von Neumann algebra

W^{*} (M, ℱ)

depends only on the class of the measure

m

(i.e. the family of all sets of

m

-measure zero), it is easy to see that this definition does not depend on the choice of

ρ

and

α

If we assume that the union of all leaves with nontrivial holonomy has measure zero, elements of

W^{*} (M, ℱ)

can be considered as measurable families

{T_{l} : l \in M / ℱ}

, where

T_{l}

is a bounded operator in

L^{2} (l)

for each leaf

l

ℱ

(see more details in [34, 33, 42] ).

A holonomy invariant measure

μ

M

defines a normal semi-finite trace

t r_{μ}

on the von Neumann algebra

W^{*} (M, ℱ)

. For any bounded measurable function

k

G

, the value

t r_{μ} (k)

is finite and is given by

t r_{μ} (k) = \int_{M} k (x) d μ (x) .

The paper [34] gives a description of weights on the von Neumann algebra

W^{*} (M, ℱ)

. As explained in [33] , the construction of [34] can be interpreted as a correspondence between weights on

W^{*} (M, ℱ)

and operator-valued densities on the leaf space

M / ℱ

Example 3.27. Consider the linear foliation

ℱ_{θ}

on the torus

T^{2}

θ \in R

is a fixed irrational number. The Lebesgue measure

μ = d x d y

T^{2}

is a holonomy invariant measure. The value of the corresponding faithful normal semi-finite trace

t r_{μ}

k \in C_{c}^{\infty} (T^{2} \times R) \subset W^{*} (T^{2}, ℱ_{θ})

is given by

t r_{μ} (k) = \int_{T^{2}} k (x, y, 0) d x d y .

A von Neumann algebra is called a factor if its center consists of operators of the form

λ \cdot I, λ \in C

Theorem 3.28. Let

(M, ℱ)

be a foliated manifold. A von Neumann algebra

W^{*} (M, ℱ)

is a factor if and only if the foliation is ergodic, that is, any bounded measurable function, constant along the leaves of the foliation

ℱ

, is constant on

M

It is known that von Neumann algebras are classified in three classes:

type I, II and III. Any von Neumann algebra

ℳ

is canonically represented as a direct sum

ℳ_{I} \oplus ℳ_{I I} \oplus ℳ_{I I I}

of von Neumann algebras, where

ℳ_{I}

ℳ_{I I}

and

ℳ_{I I I}

are von Neumann algebras of type I, II and III accordingly.

Theorem 3.29. Let

(M, ℱ)

be a foliated manifold. The von Neumann algebra

W^{*} (M, ℱ)

is of:

(1) type I if and only if the leaf space is isomorphic to the standard Borel space.
(2) type II if and only if there is a holonomy invariant transverse measure and the algebra is not of type I.
(3) type III if and only if there is no holonomy invariant transverse measure.

3.4 $C^{*}$ -modules and vector bundles

A noncommutative generalization of the notion of vector bundle (as a topological object) is the notion of Hilbert

C^{*}

-module. Hilbert

C^{*}

-modules are also natural generalizations of Hilbert spaces, which arise if we replace the field of scalars

C

to an arbitrary

C^{*}

-algebra. The theory of Hilbert

C^{*}

-modules appeared in the papers [145, 157] and has found many applications in the theory of operator algebras and its applications (see more detailed expositions of basic facts of this theory in [118, 127, 104, 183] ).

Definition 3.30. [145, 157] Let

B

be a

C^{*}

-algebra. A pre-Hilbert

B

-module is a right

B

-module

X

equipped with a sesquilinear map (linear in the second argument)

〈 \cdot, \cdot 〉_{X} : X \times X \to B

, satisfying the following conditions:

(1) $〈 x, x 〉_{X} \geq 0$ for any $x \in X$ ;
(2) $〈 x, x 〉_{X} = 0$ if and only if $x = 0$ ;
(3) $〈 y, x 〉_{X} = 〈 x, y 〉_{X}^{*}$ for any $x, y \in X$ ;
(4) $〈 x, y b 〉_{X} = 〈 x, y 〉_{X} b$ for any $x, y \in X, b \in B$ .

The map

〈 \cdot, \cdot 〉_{X}

is called a

B

-valued inner product.

Let

X

be a pre-Hilbert

B

-module. It can be shown that the formula

∥ x ∥_{X} = ∥ 〈 x, x 〉_{X} ∥^{1 / 2}

defines a norm on

X

. If

X

is complete in the norm

∥ \cdot ∥_{X}

, then

X

is called a Hilbert

C^{*}

-module. In a general case, the action of

B

and the inner product on

X

are extended to its completion

\tilde{X}

, making

\tilde{X}

into a Hilbert

C^{*}

-module.

Example 3.31. If

E

is a Hermitian vector bundle on a compact manifold

X

, then the space of its continuous sections

C (X, E)

is a Hilbert module over the algebra

C (X)

of continuous functions on

X

. The action of

C (X)

C (X, E)

is given by

(a \cdot s) (x) = a (x) s (x), x \in X,

and the inner product

〈 s_{1}, s_{2} 〉 (x) = 〈 s_{1} (x), s_{2} (x) 〉_{E_{x}}, x \in X .

Definition 3.32. A vector bundle

E

on a foliated manifold

(M, ℱ)

is called holonomy equivariant, if there is given a representation

T

of the holonomy groupoid

G

of the foliation

ℱ

in the fibers of

E

, that is, for any

γ \in G, γ : x \to y

, there is defined a linear operator

T (γ) : E_{x} \to E_{y}

such that

T (γ_{1} γ_{2}) = T (γ_{1}) T (γ_{2})

for any

γ_{1}, γ_{2} \in G

with

r (γ_{2}) = s (γ_{1})

A Hermitian vector bundle

E

on a foliated manifold

(M, ℱ)

is called holonomy equivariant, if it is a holonomy equivariant vector bundle and the representation

T

is unitary:

T (γ^{- 1}) = T (γ)^{*}

for any

γ \in G

For any holonomy equivariant vector bundle

E \to M

, the action of the groupoid

G

E

defines a horizontal foliation

ℱ_{E}

E

of the same dimension as the foliation

ℱ

. The leaf of

ℱ_{E}

through a point

v \in E

consists of all points of the form

T (γ)^{- 1} (v)

with

γ \in G

r (γ) = π (v)

. Thus, any holonomy equivariant vector bundle is foliated in the sense of [107] .

Definition 3.33. A vector bundle

p : P \to M

is called foliated, if there is a foliation

\bar{ℱ}

P

of the same dimension as

ℱ

such that its leaves are transversal to the fibers of

p

and are mapped by

p

to the leaves of

ℱ

Equivalently, one can say that a foliated vector bundle is a vector bundle

P

M

, which has a flat connection in the space

C^{\infty} (M, P)

defined along the leaves of

ℱ

, that is, an operator

\nabla : X (ℱ) \times C^{\infty} (M, P) \to C^{\infty} (M, P),

satisfying, for any

f \in C^{\infty} (M), X \in X (ℱ), s \in C^{\infty} (M, P)

, the standard conditions

\nabla_{f X} = f \nabla_{X}, \nabla_{X} (f s) = (X f) s + f \nabla_{X} s,

and also the flatness condition

[\nabla_{X}, \nabla_{Y}] = \nabla_{[X, Y]}, X, Y \in X (ℱ) .

The parallel transport along leafwise paths associated with the connection

\nabla

defines an action of the fundamental groupoid of the foliation

Π (M, ℱ)

in the fibers of the foliated vector bundle

P

. In general, the parallel transport may depend on the holonomy of the corresponding path, therefore, this action does not necessarily pull down to an action of the holonomy groupoid in the fibers of

P

, and, therefore, a foliated vector bundle is not necessarily holonomy equivariant.

Let

E

be a holonomy equivariant vector bundle. The holonomy groupoid

G_{E}

of the horizontal foliation

(E, ℱ_{E})

is described as follows:

G_{E} = r^{*} (E) = {(γ, v) \in G \times E : r (γ) = π (v)},

(G_{E})^{(0)} = E

, the source map

s : G_{E} \to E

is given by

s (γ, v) = T (γ)^{- 1} (v)

, the range map

r : G_{E} \to E

r (γ, v) = v

and the composition has the form

(γ, v) (γ^{'}, v^{'}) = (γ γ^{'}, v)

, where

v^{'} = T (γ)^{- 1} (v)

Example 3.34 (Normal bundle). The normal bundle

τ_{x} = T_{x} M / T_{x} ℱ, x \in M,

is a holonomy equivariant vector bundle, if it equipped with the action of the holonomy groupoid

G

by the linear holonomy map

d h_{γ} : τ_{x} \to τ_{y}, γ : x \to y

. The corresponding partial flat connection defined along the leaves of

ℱ

is the Bott connection

\overset{\circ}{\nabla}

(cf. ( 2.1 )). The normal bundle

τ

is a holonomy equivariant Hermitian vector bundle, if

ℱ

is a Riemannian foliation.

Example 3.35 (Transverse cotangent bundle). The dual example to the previous one is given by the conormal bundle

N^{*} ℱ

equipped with the action of the holonomy groupoid

G

by the linear holonomy map

(d h_{γ}^{*})^{- 1} : N_{x}^{*} ℱ \to N_{y}^{*} ℱ

for

γ : x \to y

. The corresponding flat connection defined along the leaves of

ℱ

is the connection

(\overset{\circ}{\nabla})^{*}

, dual to the Bott connection (cf. ( 2.3 )).

The associated groupoid, which will be denoted by

G_{ℱ_{N}}

, is described as follows:

G_{ℱ_{N}} = {(γ, η) \in G \times N^{*} ℱ : r (γ) = π (η)}

with the source map

s : G_{ℱ_{N}} \to N^{*} ℱ, s (γ, η) = d h_{γ}^{*} (η)

, the range map

r : G_{ℱ_{N}} \to N^{*} ℱ, r (γ, η) = η

and the composition

(γ, η) (γ^{'}, η^{'}) = (γ γ^{'}, η)

defined in the case when

η^{'} = d h_{γ}^{*} (η)

. The projection

π : N^{*} ℱ \to M

induces a map

π_{G} : G_{ℱ_{N}} \to G_{ℱ}

by the formula

π_{G} (γ, η) = γ, (γ, η) \in G_{ℱ_{N}} .

The corresponding horizontal foliation coincides with the linearized foliation

ℱ_{N}

N^{*} ℱ

(see the symplectic description of this foliation in Example 2.46 ), which can be described in the following way (cf.

also an invariant definition in [136] ). The leaf

{\tilde{L}}_{η}

ℱ_{N}

through a point

η \in N^{*} ℱ

consists of all points of the form

d h_{γ}^{*} (η)

with

γ \in G

such that

r (γ) = π (η)

. It is diffeomorphic to the holonomy covering

G^{x}

of the leaf

L_{x}, x = π (η)

ℱ

through

x

. Each leaf has trivial holonomy, and the holonomy groupoid of the linearized foliation

ℱ_{N}

coincides with

G_{ℱ_{N}}

Let us also note that the canonical relation

Γ

T^{*} M

associated with

N^{*} ℱ

(cf. Example 2.46 ) is given by the (one-to-one) immersion

(r, s) : G_{ℱ_{N}} \to T^{*} M \times T^{*} M .

For any holonomy equivariant vector bundle

E

on a foliated manifold

(M, ℱ)

, there is defined a pre-Hilbert

C_{c}^{\infty} (G)

-module

ℰ_{\infty}

. As a linear space,

ℰ_{\infty}

coincides with

C_{c}^{\infty} (G, r^{*} E)

. The module structure on

ℰ_{\infty}

is introduced in the following way: the action of

f \in C_{c}^{\infty} (G)

s \in ℰ_{\infty}

is given by

(s * f) (γ) = \int_{G^{y}} s (γ^{'}) f ({γ^{'}}^{- 1} γ) d ν^{y} (γ^{'}), γ \in G^{y},

and the inner product on

ℰ_{\infty}

with values in

C_{c}^{\infty} (G)

is given by

〈 s_{1}, s_{2} 〉 (γ) = \int_{G^{y}} 〈 s_{1} ({γ^{'}}^{- 1}), s_{2} ({γ^{'}}^{- 1} γ) 〉_{E_{s (γ^{'})}} d ν^{y} (γ^{'}), s_{1}, s_{2} \in ℰ_{\infty} .

The completion of

ℰ_{\infty}

in the norm

∥ s ∥ = ∥ R (〈 s, s 〉) ∥^{1 / 2}

defines a

C^{*}

-Hilbert module over

C_{r}^{*} (G)

, that we denote by

ℰ

. It is equipped with a

C_{r}^{*} (G)

-sesquilinear form

〈 \cdot, \cdot 〉

with values in

C_{r}^{*} (G)

, which is the extension by continuity of the sesquilinear form on

ℰ_{\infty}

. The

C^{*}

-Hilbert module

ℰ

can be considered as a noncommutative analogue of the algebra of continuous sections of the bundle

E

considered as a bundle on the leaf space

M / ℱ

There is also a left action of

C_{c}^{\infty} (G)

ℰ_{\infty}

given by

\begin{matrix} (f * s) (γ) = \int_{G^{y}} f (γ^{'}) T (γ^{'}) [s ({γ^{'}}^{- 1} γ)] d ν^{y} (γ^{'}), γ \in G^{y}, \end{matrix}

(3.4)

where

f \in C_{c}^{\infty} (G)

and

s \in ℰ_{\infty}

. Unlike the right action, the left action does not extend to an action of

C_{r}^{*} (G)

by bounded endomorphisms of the

C^{*}

-Hilbert module

ℰ

over

C_{r}^{*} (G)

. As shown in [42] (see also [38] ), for any

f \in C_{c}^{\infty} (G)

the formula ( 3.4 ) defines an endomorphism

λ (f)

of the

C^{*}

-Hilbert module

ℰ

with the adjoint, given by

(λ (f)^{*} s) (γ) = \int_{G^{y}} f^{♯} (γ^{'}) T (γ^{'}) [s ({γ^{'}}^{- 1} γ)] d ν^{y} (γ^{'}), γ \in G^{y},

where

f^{♯} (γ) = \bar{f} (γ^{- 1}) Δ (γ)

and

Δ (γ) : E_{r (γ)} \to E_{r (γ)}

is the linear operator given by

Δ (γ) = (T (γ)^{- 1})^{*} T (γ)^{- 1} .

Thus, if the Hermitian structure on

E

is not holonomy invariant, the representation

λ

is unbounded with respect to the

C^{*}

-norm on

C_{r}^{*} (G)

Nevertheless, it can be shown that the homomorphism

λ

is a densely defined closable homomorphism of

C^{*}

-algebras. Hence, the domain of the closure

\bar{λ}

of the homomorphism

λ

endowed with the graph norm

∥ x ∥_{λ} = ∥ x ∥ + ∥ λ (x) ∥

is a Banach algebra

B

, dense in the

C^{*}

-algebra

C_{r}^{*} (G)

. Thus, the

C^{*}

-Hilbert module

ℰ

is a

B

C_{r}^{*} (G)

-bimodule.

In particular, if we take

E

to be the holonomy equivariant bundle

\land^{j} N^{*} ℱ

, we get the

C^{*}

-Hilbert module of transverse differential

j

-forms

Ω^{j}

over the

C^{*}

-algebra

C_{r}^{*} (G)

as the completion of the pre-Hilbert

C_{c}^{\infty} (G, | T G |^{1 / 2})

-module

Ω_{\infty}^{j} = C_{c}^{\infty} (G, r^{*} \land^{j} N^{*} ℱ \otimes | T G |^{1 / 2})

. There is a product

Ω_{\infty}^{j} \times Ω_{\infty}^{k} \overset{\land}{\to} Ω_{\infty}^{j + k}

, compatible with the bimodule structure:

\begin{matrix} (ω \land ω_{1}) (γ) = \int_{G^{y}} ω (γ_{1}) \land γ_{1} ω_{1} (γ_{1}^{- 1} γ), ω \in Ω_{\infty}^{j}, ω_{1} \in Ω_{\infty}^{k} . \end{matrix}

(3.5)

For any holonomy equivariant vector bundle

E

on a foliated manifold

(M, ℱ)

, there is defined a natural representation of

C_{c}^{\infty} (G)

in the space of sections of this bundle. It is obtained as the restriction of the representation

R_{E}

of the algebra

C_{c}^{\infty} (G, ℒ (E))

C_{c}^{\infty} (G)

, which is embedded in

C_{c}^{\infty} (G, ℒ (E))

by means of the homomorphism

k \in C_{c}^{\infty} (G) \mapsto k T \in C_{c}^{\infty} (G, ℒ (E)) .

{ν^{x} : x \in G^{(0)}}

is a smooth Haar system on

G

, then this representation, which we will also denote by

R_{E}

, is defined as follows. For any

u \in C^{\infty} (M, E)

, the section

R_{E} (k) u \in C^{\infty} (M, E)

is given by

\begin{matrix} R_{E} (k) u (x) = \int_{G^{x}} k (γ) T (γ) [u (s (γ))] d ν^{x} (γ), x \in M . \end{matrix}

(3.6)

E

is a holonomy equivariant Hermitian vector bundle on

M

, then

R_{E}

is a

*

-representation. In this case, we define a

C^{*}

-algebra

C_{E}^{*} (G)

as the closure of

R_{E} (C_{c}^{\infty} (G))

in the uniform topology of

ℒ (L^{2} (M, E))

By Theorem 2.1 in [68] , there is an estimate

\begin{matrix} ∥ k ∥_{C_{r}^{*} (G, E)} \leq ∥ R_{E} (k) ∥, k \in C_{c}^{\infty} (G, ℒ (E)), \end{matrix}

(3.7)

therefore, the reduced

C^{*}

-algebra

C_{r}^{*} (G, E)

is the quotient of the algebra

C_{E}^{*} (G)

, and there is a natural projection

π : C_{E}^{*} (G) \to C_{r}^{*} (G, E)

4 Noncommutative topology

4.1 $K$ -theory and $K$ -homology

One of the main tools for the investigation of topological spaces is the topological

K

-theory (see [1, 108, 132] ). Recall that the group

K^{0} (X)

associated with a compact topological space

X

is generated by stable equivalence classes of locally trivial finite-dimensional complex vector bundles on

X

Equivalently, elements of

K^{0} (X)

are described as stable isomorphism classes of finitely generated projective modules over the algebra

C (X)

or as equivalence classes of projections in the matrix algebra over

C (X)

(see below).

In this Section, we briefly recall the definition of the

K

-theory for

C^{*}

-algebras, the noncommutative analogue of the topological

K

-theory, and also the definition of the

K

-homology groups (see, for instance, [19, 42, 98, 108, 160, 183] for further information).

Let

A

be a unital

C^{*}

-algebra. Denote by

M_{n} (A)

the algebra of

n \times n

-matrices with elements from

A

. We will assume that

M_{n} (A)

is embedded into

M_{n + 1} (A)

by the map

X \to (\begin{matrix} X & 0 \\ 0 & 0 \end{matrix})

. Let

M_{\infty} (A) = lim \to M_{n} (A)

The group

K_{0} (A)

is defined as the set of homotopy equivalence classes of projections (

p^{2} = p = p^{*}

) in

M_{\infty} (A)

equipped with the direct sum operation

p_{1} \oplus p_{2} = (\begin{matrix} p_{1} & 0 \\ 0 & p_{2} \end{matrix}) .

Denote by

G L_{n} (A)

the group of invertible

n \times n

-matrices with elements from

A

. We will assume that

G L_{n} (A)

is embedded into

G L_{n + 1} (A)

by the map

X \to (\begin{matrix} X & 0 \\ 0 & 1 \end{matrix})

. Let

G L_{\infty} (A) = lim \to G L_{n} (A)

. The group

K_{1} (A)

is defined as the set of homotopy equivalence classes of unitary matrices (

u^{*} u = u u^{*} = 1

) in

G L_{\infty} (A)

, equipped with the direct sum operation.

A

has no unit and

A^{+}

is the algebra, obtained by adding the unit to the algebra

A

, then we have the homomorphism

i : C \to A^{+} : λ \mapsto λ \cdot 1

, inducing a homomorphism

i_{*} : K_{0} (C) \to K_{0} (A^{+})

, and

K_{0} (A)

is defined as the kernel of this homomorphism. Moreover, by definition,

K_{1} (A) = K_{1} (A^{+})

For an arbitrary algebra

A

over

C

, there are defined the groups

K_{0} (A)

and

K_{1} (A)

of the algebraic

K

-theory (see, for instance, [7, 130] ). The group

K_{0} (A)

is defined similarly to the group

K_{0} (A)

of the topological

K

-theory, using idempotents (

e^{2} = e

) in

M_{\infty} (A)

instead of projections. The group

K_{1} (A)

is defined as the quotient of the group

G L_{\infty} (A)

by the commutant

[G L_{\infty} (A), G L_{\infty} (A)]

The construction of the

K

-homology for noncommutative algebras is based on the notion of Fredholm module. This notion is a functional-analytic abstraction of the notion of elliptic pseudodifferential operator on a compact manifold. The idea that abstract elliptic operators can be naturally considered as elements of the

K

-homology groups, suggested by Atiyah [2] , was realized by Kasparov (see [110] ). Recall that a Hilbert space

H

is called

Z_{2}

-graded, if there is given its decomposition into a direct sum of Hilbert subspaces

H = H_{0} \oplus H_{1}

. Equivalently, a

Z_{2}

-grading on

H

is defined by a self-adjoint operator

γ \in ℒ (H)

such that

γ^{2} = 1

. Given the decomposition

H = H^{+} \oplus H^{-}

, the operator

γ

has the matrix

(\begin{matrix} 1 & 0 \\ 0 & - 1 \end{matrix}) .

Definition 4.1. A Fredholm module (or a

K

-cycle) over a

C^{*}

-algebra

A

is a pair

(H, F)

, where

(1) $H$ is a Hilbert space, equipped with a $*$ -representation $ρ$ of the algebra $A$ ;
(2) $F$ is a bounded operator in $H$ such that, for any $a \in A$ , the operators $(F^{2} - 1) ρ (a)$ , $(F - F^{*}) ρ (a)$ and $[F, ρ (a)]$ are compact in $H$ .

A Fredholm module

(H, F)

is called even, if the Hilbert space

H

is endowed with a

Z_{2}

-grading

γ

, the operators

ρ (a)

are even,

γ ρ (a) = ρ (a) γ

, and the operator

F

is odd,

γ F = - F γ

. In the opposite case, it is called odd.

Recall that, for any

p \geq 1

, the Schatten class

ℒ^{p} (ℋ)

consists of all compact operators

T

in a Hilbert space

ℋ

such that the operator

| T |^{p}

is a trace class operator. Let

μ_{1} (T) \geq μ_{2} (T) \geq \dots

be the characteristic numbers (

s

-numbers) of a compact operator

T

ℋ

, that is, the eigenvalues of the operator

| T | = \sqrt{T^{*} T}

, taken with multiplicities. Then

T \in ℒ^{p} (ℋ) \Leftrightarrow t r | T |^{p} = \sum_{n = 1}^{\infty} | μ_{n} (T) |^{p} < \infty .

Definition 4.2. A Fredholm module

(H, F)

is called

p

-summable, if

(F^{2} - 1) ρ (a)

(F - F^{*}) ρ (a)

and

[F, ρ (a)]

belong to

ℒ^{p} (H)

The homology groups

K^{0} (A)

(

K^{1} (A)

) are defined as the sets of homotopy equivalence classes of even (resp. odd) Fredholm modules over

A

. The direct sum operation defines the structure of an abelian group on

K^{0} (A)

and

K^{1} (A)

Example 4.3 ([39] ). Let

M

be a compact manifold,

E

a Hermitian vector bundle on

M

. Then the pair

(H, F)

, where:

$∙$ $H = L^{2} (M, E)$ ;
$∙$ $F \in Ψ^{0} (M, E)$ is an elliptic operator with the principal symbol $σ_{F}$ such that $σ_{F}^{2} = 1, σ_{F}^{*} = σ_{F}$ (for instance, one can take an operator of the form $D (1 + D^{2})^{- 1 / 2}$ with some self-adjoint elliptic operator $D \in Ψ^{m} (M, E), m > 0$ )

is a Fredholm module over

C^{\infty} (M)

Example 4.4 ([39] ). Let

(M, ℱ)

be a compact foliated manifold,

E

a holonomy equivariant Hermitian vector bundle on

M

. Then the pair

(H, F)

, where:

$∙$ $H = L^{2} (M, E)$ ;
$∙$ $F \in Ψ^{0} (M, E)$ is a transversally elliptic operator with the holonomy invariant (see Definition 7.7 ) transversal principal symbol $σ_{F}$ such that $σ_{F}^{2} = 1, σ_{F}^{*} = σ_{F}$ .

is a Fredholm module over

C_{c}^{\infty} (G)

A Fredholm module

(H, F)

defines an index map

ind : K_{*} (A) \to Z

In the even case, with respect to the decomposition

H = H^{+} \oplus H^{-}

given by the

Z_{2}

-grading of

H

, the operator

F

takes the form

\begin{matrix} F = (\begin{matrix} 0 & F^{-} \\ F^{+} & 0 \end{matrix}), F_{\pm} : H^{\mp} \to H^{\pm} . \end{matrix}

(4.1)

For any projection

e \in M_{q} (A)

, the operator

e (F^{+} \otimes 1) e

, acting from

e (H^{+} \otimes C^{q})

e (H^{-} \otimes C^{q})

, is Fredholm, its index depends only on the homotopy class of

e

. Therefore, a map

ind : K_{0} (A) \to Z

is defined as

\begin{matrix} ind [e] = ind e (F^{+} \otimes 1) e . \end{matrix}

(4.2)

In the odd case, for a unitary matrix

U \in G L_{q} (A)

, the operator

(P \otimes 1) U (P \otimes 1)

, where

P = \frac{1 + F}{2}

, is a Fredholm operator. Moreover, the index of the operator

(P \otimes 1) U (P \otimes 1)

depends only on the homotopy class of

U

Therefore, the map

ind : K_{1} (A) \to Z

is given by

\begin{matrix} ind [U] = ind (P \otimes 1) U (P \otimes 1) . \end{matrix}

(4.3)

In both cases, both in the even one, and in the odd one, the map

ind

depends only on the class defined by the Fredholm module

(H, F)

in the

K

-homology group.

Computation of the

K

-theory for the

C^{*}

-algebra of a foliation is not an easy problem. In the case of the linear foliation on

T^{2}

, the computation of the

K

-theory of the

C^{*}

-algebra of the foliation can be reduced to the computation of the

K

-theory of the

C^{*}

-algebra

A_{θ}

(see Example 4.10 ) and was given in [150] . For arbitrary foliations on

T^{2}

and

S^{3}

, the computation of the

K

-theory of the

C^{*}

-algebra of the foliation was done in [178] .

In [8] (see also [55] ) a geometrical construction of elements of the group

K (C^{*} (M, ℱ))

was suggested. For any topological groupoid

C

, one can define the classifying space

B C

, which is constructed, using a modification of the classical Milnor's construction of the classifying space of a group [89, 129] . In particular, the classifying space

B G

of the holonomy groupoid of a foliation is defined (see [37,Chapter9] ). The normal bundle

τ

M

defines a vector bundle on

B G

, which is also denoted by

τ

. The geometric group

K_{*, τ} (B G)

is defined as the relative

K

-theory group

K_{*, τ} (B G) = K_{*} (B (τ), S (τ)),

where

B (τ)

is the unit ball subbundle of the bundle

τ

and

S (τ)

is the unit sphere bundle. In [8] (cf. also [55] ), a map (the topological index)

μ : K_{*, τ} (B G) \to K (C^{*} (M, ℱ))

was constructed by means of topological constructions. The Baum-Connes conjecture claims that

μ

is an isomorphism. We refer the reader to the papers [93, 124, 125, 142, 143, 170, 171, 172, 178, 182, 181] for various aspects of the Baum-Connes conjecture for foliations and related computations of the

K

-theory for

C^{*}

-algebras of foliations.

4.2 Strong Morita equivalence

The notion of isometric

*

-isomorphism between

C^{*}

-algebras is a natural analogue of the notion of homeomorphism of topological spaces. However, there is a wider equivalence relation for

C^{*}

-algebras, strong Morita equivalence, introduced by Rieffel [157, 159] , which preserves many invariants of

C^{*}

-algebras, for instance, the K-theory, the space of irreducible representations, the cyclic cohomology, and coincides with the isomorphism on the class of commutative

C^{*}

-algebras. The isometric

*

-isomorphism of

C^{*}

-algebras of foliations is related with the isomorphism of the corresponding holonomy groupoids, while their strong Morita equivalence is related with the isomorphism of the corresponding leaf spaces. Therefore, in many respects the notion of strong Morita equivalence is a more adequate notion for the study of transverse geometric structures on foliated manifolds.

Definition 4.5. Let

A

and

B

C^{*}

-algebras. An

A

B

-equivalence bimodule is an

A

B

-bimodule

X

, endowed with

A

-valued and

B

-valued inner products

〈 \cdot, \cdot 〉_{A}

and

〈 \cdot, \cdot 〉_{B}

accordingly, such that

X

is a right Hilbert

B

-module and a left Hilbert

A

-module with respect to these inner products, and, moreover,

(1) $〈 x, y 〉_{A} z = x 〈 y, z 〉_{B}$ for any $x, y, z \in X$ ;
(2) The set $〈 X, X 〉_{A}$ generates a dense subset in $A$ , and the set $〈 X, X 〉_{B}$ generates a dense subset in $B$ .

We call algebras

A

and

B

strongly Morita equivalent, if there is an

A

B

-equivalence bimodule.

It is not difficult to show that the strong Morita equivalence is an equivalence relation.

For any linear space

L

, denote by

\tilde{L}

the conjugate complex linear space, which coincides with

L

as a set and has the same addition operation, but the multiplication by scalars is given by the formula

λ \tilde{x} = (\bar{λ} x)^{\sim}

. If

X

is an

A

B

-equivalence bimodule, then

\tilde{X}

is endowed with the structure of a

B

A

-equivalence bimodule. For instance,

b \tilde{x} a = (a^{*} x b^{*})^{\sim} .

[]

X

A

B

E \to X \otimes_{B} E

B

A

F \to F \otimes_{B} \tilde{X}

Yuri A. Kordyukov