K-theory
Topological K-theory is a generalised cohomology theory on the category
of compact
Hausdorff spaces.
It classifies the vector bundles
over a space X up to stable equivalences.
Equivalently, via the Serre-Swan theorem, it classifies the finitely generated projective modules over the C*-algebra
C(X).
Let A be a unital C*-algebra over ℂ and denote by M∞(A) the algebraic direct limit of matrix algebras Mn(A) under the embeddings
Mn(A)→Mn+1(A):a↦(a000).
Identify the completion of M∞(A) with the stable algebra A⊗𝕂 (where 𝕂 is the compact operators
on l2(ℕ)),
which we will continue to denote by M∞(A).
The K0(A) group is the Grothendieck group (abelian group
of formal differences
) of the homotopy classes of the projections in M∞(A).
Two projections p and q are homotopic
if there exists a norm continuous
path of projections from p to q.
Let p∈Mm(A) and q∈Mn(A) be two projections.
The sum of their homotopy classes [p] and [q] is the homotopy class of their direct sum
:
[p]+[q]=[p⊕q] where p⊕q=diag(p,q)∈Mm+n(A).
Alternatively, one can consider equivalence classes
of projections up to unitary transformations.
Unitary equivalence coincides with homotopy equivalence
in M∞(A) (or Mn(A) for n large enough).
Denote by U∞(A) the direct limit of unitary groups Un(A) under the embeddings
Un(A)→Un+1(A):u↦(u001).
Give U∞(A) the direct limit topology, i.e. a subset U of U∞(A) is open if and only if
U∩Un(A) is an open subset of Un(A), for all n.
The K1(A) group is the Grothendieck group (abelian group of formal differences) of the homotopy classes of the unitaries in U∞(A).
Two unitaries u and v are homotopic if there exists a norm continuous path of unitaries from u to v.
Let u∈Um(A) and v∈Un(A) be two unitaries.
The sum of their homotopy classes [u] and [v] is the homotopy class of their direct sum:
[u]+[v]=[u⊕v] where u⊕v=diag(u,v)∈Um+n(A).
Equivalently, one can work with invertibles in GL∞(A)
(an invertible g is connected to the unitary u=g|g|-1 via the homotopy
t→g|g|-t).
Higher K-groups can be defined through repeated suspensions,
Kn(A)=K0(SnA). | (1) |
But, the Bott periodicity theorem means that
K1(SA)≅K0(A). | (2) |
The main properties of Ki are:
Ki(A⊕B) | = | Ki(A)⊕Ki(B), | (3) | ||
Ki(Mn(A)) | = | Ki(A) | (4) | ||
(5) | |||||
(6) |
There are three flavours of topological K-theory to handle the cases of being complex (over ), real (over ) or Real (with a given real structure).
(7) | |||||
(8) | |||||
(9) |
Real K-theory has a Bott period of 8, rather than 2.
References
- 1 N. E. Wegge-Olsen, K-theory and -algebras. Oxford science publications. Oxford University Press, 1993.
- 2 B. Blackadar, K-Theory for Operator Algebras. Cambridge University Press, 2nd ed., 1998.
- 3 M. Rørdam, F. Larsen and N. J. Laustsen, An Introduction to K-Theory for -Algebras. Cambridge University Press, 2000.
Title | K-theory |
---|---|
Canonical name | Ktheory |
Date of creation | 2013-03-22 12:58:06 |
Last modified on | 2013-03-22 12:58:06 |
Owner | mhale (572) |
Last modified by | mhale (572) |
Numerical id | 17 |
Author | mhale (572) |
Entry type | Topic |
Classification | msc 19-00 |
Synonym | Topological K-theory |
Related topic | KHomology |
Related topic | AlgebraicKTheory |
Related topic | GrothendieckGroup |