In the
mathematical
Mathematics is an area of knowledge that includes the topics of numbers, formulas and related structures, shapes and the spaces in which they are contained, and quantities and their changes. These topics are represented in modern mathematics ...
field of
functional analysis
Functional analysis is a branch of mathematical analysis, the core of which is formed by the study of vector spaces endowed with some kind of limit-related structure (e.g. inner product, norm, topology, etc.) and the linear functions defined ...
there are several standard
topologies which are given to the algebra of
bounded linear operators on a
Banach space
In mathematics, more specifically in functional analysis, a Banach space (pronounced ) is a complete normed vector space. Thus, a Banach space is a vector space with a metric that allows the computation of vector length and distance between ve ...
.
Introduction
Let
be a sequence of linear operators on the Banach space . Consider the statement that
converges to some operator on .
This could have several different meanings:
* If
, that is, the
operator norm of
(the supremum of
, where ranges over the
unit ball
Unit may refer to:
Arts and entertainment
* UNIT, a fictional military organization in the science fiction television series ''Doctor Who''
* Unit of action, a discrete piece of action (or beat) in a theatrical presentation
Music
* ''Unit'' (al ...
in ) converges to 0, we say that
in the
uniform operator topology.
* If
for all
, then we say
in the
strong operator topology.
* Finally, suppose that for all we have
in the
weak topology
In mathematics, weak topology is an alternative term for certain initial topologies, often on topological vector spaces or spaces of linear operators, for instance on a Hilbert space. The term is most commonly used for the initial topology of a ...
of . This means that
for all
linear functionals on . In this case we say that
in the
weak operator topology.
List of topologies on B(''H'')
There are many topologies that can be defined on besides the ones used above; most are at first only defined when is a Hilbert space, even though in many cases there are appropriate generalisations.
The topologies listed below are all locally convex, which implies that they are defined by a family of
seminorms.
In analysis, a topology is called strong if it has many open sets and weak if it has few open sets, so that the corresponding modes of convergence are, respectively, strong and weak.
(In topology proper, these terms can suggest the opposite meaning, so strong and weak are replaced with, respectively, fine and coarse.)
The diagram on the right is a summary of the relations, with the arrows pointing from strong to weak.
If is a Hilbert space, the
Hilbert space
In mathematics, Hilbert spaces (named after David Hilbert) allow generalizing the methods of linear algebra and calculus from (finite-dimensional) Euclidean vector spaces to spaces that may be infinite-dimensional. Hilbert spaces arise natu ...
has a (unique)
predual ,
consisting of the trace class operators, whose dual is .
The seminorm for ''w'' positive in the predual is defined to be
.
If is a vector space of linear maps on the vector space , then is defined to be the weakest topology on such that all elements of are continuous.
* The
norm topology
In mathematics, the operator norm measures the "size" of certain linear operators by assigning each a real number called its . Formally, it is a norm defined on the space of bounded linear operators between two given normed vector spaces.
Introdu ...
or uniform topology or uniform operator topology is defined by the usual norm , , ''x'', , on . It is stronger than all the other topologies below.
* The
weak (Banach space) topology is , in other words the weakest topology such that all elements of the dual are continuous. It is the weak topology on the Banach space . It is stronger than the ultraweak and weak operator topologies. (Warning: the weak Banach space topology and the weak operator topology and the ultraweak topology are all sometimes called the weak topology, but they are different.)
* The
Mackey topology or Arens-Mackey topology is the strongest locally convex topology on such that the dual is , and is also the uniform convergence topology on , -compact convex subsets of . It is stronger than all topologies below.
* The σ-strong-
* topology or ultrastrong-
* topology is the weakest topology stronger than the ultrastrong topology such that the adjoint map is continuous. It is defined by the family of seminorms and for positive elements of . It is stronger than all topologies below.
*The σ-strong topology or
ultrastrong topology or strongest topology or strongest operator topology is defined by the family of seminorms for positive elements of . It is stronger than all the topologies below other than the strong
* topology. Warning: in spite of the name "strongest topology", it is weaker than the norm topology.)
*The σ-weak topology or ultraweak topology or
weak-* operator topology or weak-* topology or weak topology or ) topology is defined by the family of seminorms , (''w'', ''x''), for elements ''w'' of . It is stronger than the weak operator topology. (Warning: the weak Banach space topology and the weak operator topology and the ultraweak topology are all sometimes called the weak topology, but they are different.)
* The
strong-* operator topology or strong-
* topology is defined by the seminorms , , ''x''(''h''), , and , , ''x''
*(''h''), , for . It is stronger than the strong and weak operator topologies.
* The
strong operator topology (SOT) or strong topology is defined by the seminorms , , ''x''(''h''), , for . It is stronger than the weak operator topology.
* The
weak operator topology (WOT) or weak topology is defined by the seminorms , (''x''(''h''
1), ''h''
2), for . (Warning: the weak Banach space topology, the weak operator topology, and the ultraweak topology are all sometimes called the weak topology, but they are different.)
Relations between the topologies
The continuous linear functionals on for the weak, strong, and strong
* (operator) topologies are the same, and are the finite linear combinations of the linear functionals
(x''h''
1, ''h''
2) for .
The continuous linear functionals on for the ultraweak, ultrastrong, ultrastrong
* and Arens-Mackey topologies are the same, and are the elements of the predual .
By definition, the continuous linear functionals in the norm topology are the same as those in the weak Banach space topology.
This dual is a rather large space with many pathological elements.
On norm bounded sets of , the weak (operator) and ultraweak topologies coincide. This can be seen via, for instance, the
Banach–Alaoglu theorem.
For essentially the same reason, the ultrastrong
topology is the same as the strong topology on any (norm) bounded subset of .
Same is true for the Arens-Mackey topology, the ultrastrong
*, and the strong
* topology.
In locally convex spaces, closure of convex sets can be characterized by the continuous linear functionals. Therefore, for a
convex subset of , the conditions that be closed in the ultrastrong
*, ultrastrong, and ultraweak topologies are all equivalent and are also equivalent to the conditions that
for all , has closed intersection with the closed ball of radius in the strong
*, strong, or weak (operator) topologies.
The norm topology is metrizable and the others are not; in fact they fail to be
first-countable
In topology, a branch of mathematics, a first-countable space is a topological space satisfying the "first axiom of countability". Specifically, a space X is said to be first-countable if each point has a countable neighbourhood basis (local bas ...
.
However, when is separable, all the topologies above are metrizable when restricted to the unit ball (or to any norm-bounded subset).
Topology to use
The most commonly used topologies are the norm, strong, and weak operator topologies.
The weak operator topology is useful for compactness arguments, because the unit ball is compact by the
Banach–Alaoglu theorem.
The norm topology is fundamental because it makes into a Banach space, but it is too strong for many purposes; for example, is not separable in this topology.
The strong operator topology could be the most commonly used.
The ultraweak and ultrastrong topologies are better-behaved than the weak and strong operator topologies, but their definitions are more complicated, so they are usually not used unless their better properties are really needed.
For example, the dual space of in the weak or strong operator topology is too small to have much analytic content.
The adjoint map is not continuous in the strong operator and ultrastrong topologies, while the strong* and ultrastrong* topologies are modifications so that the adjoint becomes continuous. They are not used very often.
The Arens–Mackey topology and the weak Banach space topology are relatively rarely used.
To summarize, the three essential topologies on are the norm, ultrastrong, and ultraweak topologies.
The weak and strong operator topologies are widely used as convenient approximations to the ultraweak and ultrastrong topologies. The other topologies are relatively obscure.
See also
*
*
*
*
*
*
*
*
*
References
* ''Functional analysis'', by Reed and Simon,
* ''Theory of Operator Algebras I'', by M. Takesaki (especially chapter II.2)
{{Topological vector spaces
Functional analysis
Topological vector spaces
*