14.2 Operator spaces and completely bounded maps

Operator spaces and completely bounded maps are key concepts in advanced C*-algebra theory. They extend the notion of normed spaces to matrices, allowing for a richer structure that captures the essence of operator algebras.

This topic dives into matrix norms, Ruan's theorem, and completely bounded/positive maps. It explores how these concepts relate to C*-algebras and their applications in quantum information theory and operator algebra theory.

Operator Spaces and Matrix Norms

Fundamental Concepts of Operator Spaces

• Operator space defines a vector space V equipped with a sequence of matrix norms satisfying specific conditions
• Matrix norm assigns a norm to matrices of any size with entries from a given vector space
• Minimal operator space structure arises from embedding V into B(H) for some Hilbert space H
• Maximal operator space structure emerges from considering all possible C*-algebra representations of V
• Ruan's theorem characterizes abstract operator spaces through a set of axioms for matrix norms
  • Establishes isometric equivalence between abstract operator spaces and concrete operator spaces

Properties and Applications of Matrix Norms

• Matrix norms generalize vector norms to matrices, measuring size or magnitude
• Different types of matrix norms include Frobenius norm, spectral norm, and induced norms
• Matrix norms satisfy properties such as non-negativity, positive scalability, and the triangle inequality
• Applications of matrix norms include error analysis, numerical stability, and convergence studies in linear algebra

Ruan's Theorem and Its Implications

• Ruan's theorem provides a crucial bridge between abstract and concrete operator spaces
• States that any abstract operator space can be realized as a subspace of B(H) for some Hilbert space H
• Implies that every operator space has a faithful representation as bounded operators on a Hilbert space
• Allows for the study of operator spaces using both abstract and concrete approaches
• Facilitates the development of operator space theory by connecting it to well-established operator algebra concepts

Completely Bounded and Positive Maps

Completely Bounded Maps and Their Properties

• Completely bounded map φ: V → W between operator spaces maintains boundedness when tensored with matrix algebras
• CB-norm of a map φ defined as the supremum of the norms of φ⊗In over all n
• Completely bounded maps form a natural class of morphisms in the category of operator spaces
• Properties of completely bounded maps include:
  • Composition of completely bounded maps remains completely bounded
  • Adjoint of a completely bounded map between C*-algebras remains completely bounded
  • Tensor product of completely bounded maps remains completely bounded

Completely Positive Maps and Their Significance

• Completely positive map φ: A → B between C*-algebras preserves positivity when tensored with matrix algebras
• Characterized by the property that [φ(aij)] remains positive for any positive matrix [aij] with entries from A
• Stinespring's dilation theorem provides a fundamental representation for completely positive maps
• Applications of completely positive maps include:
  • Quantum channels in quantum information theory
  • State transformations in quantum mechanics
  • Conditional expectations in operator algebra theory

Operator Systems and Their Role

• Operator system defines a self-adjoint subspace of a C*-algebra containing the identity
• Serves as a natural domain for completely positive maps
• Characterized by the existence of a matrix ordering and an order unit
• Archimedean property ensures that small positive elements remain positive under scalar multiplication
• Examples of operator systems include:
  • The space of n×n matrices
  • The self-adjoint part of a C*-algebra
  • The dual space of a C*-algebra

Advanced Topics in Operator Spaces

Haagerup Tensor Product and Its Properties

• Haagerup tensor product ⊗h provides a way to construct new operator spaces from existing ones
• Defined for operator spaces U and V as the completion of U ⊗ V with respect to a specific norm
• Properties of the Haagerup tensor product include:
  • Associativity: (U ⊗h V) ⊗h W ≅ U ⊗h (V ⊗h W)
  • Commutativity up to complete isometry: U ⊗h V ≅ V ⊗h U
  • Injectivity for subspaces: If U1 ⊆ U2 and V1 ⊆ V2, then U1 ⊗h V1 ⊆ U2 ⊗h V2
• Applications of the Haagerup tensor product include:
  • Study of completely bounded multilinear maps
  • Construction of operator space analogues of classical Banach space tensor products

Injective Envelope and Its Significance

• Injective envelope I(X) of an operator space X represents the smallest injective operator space containing X
• Characterized by the rigidity property: any completely contractive map φ: I(X) → I(X) fixing X must be the identity
• Construction of the injective envelope involves:
  • Embedding X into a large injective operator space (B(H))
  • Applying a sequence of completely contractive projections
• Properties of the injective envelope include:
  • Uniqueness up to complete isometry
  • Preservation of C*-algebraic structure when X is a C*-algebra
• Applications of the injective envelope include:
  • Extension problems for completely bounded maps
  • Study of amenability and injectivity in operator algebras
  • Construction of universal objects in operator space theory