Abstract:By the Choi matrix criteria it is easy to determine if a specific linear matrix map is completely positive, but to establish whether a linear matrix map is positive is much less straightforward. In this paper we consider classes of linear matrix maps, determined by structural conditions on an associated matrix, for which positivity and complete positivity coincide. The basis of our proofs lies in a representation of $*$-linear matrix maps going back to work of R.D. Hill which enables us to formulate a sufficient condition in terms of surjectivity of certain bilinear maps.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about when the positivity and complete positivity in linear matrix mappings are the same. Specifically, the author focuses on whether the positivity and complete positivity of linear matrix mappings can coincide under certain structural conditions. This involves how to determine whether a linear matrix mapping is positive or completely positive through structural conditions.
### Background and Problem Description
In many fields of mathematics and mathematical physics, positive matrix mappings and completely positive matrix mappings play important roles. However, it is relatively simple to verify whether a specific linear matrix mapping is completely positive (through the Choi matrix criterion), while verifying whether it is positive is much more complicated. Therefore, this paper aims to find a class of linear matrix mappings such that under their structural conditions, the positivity and complete positivity can be the same.
### Main Research Contents
1. **Differences between Positivity and Complete Positivity**:
- Positivity: For all \( x \in F^n \) and \( z \in F^q \), satisfy \( (z \otimes x)^* L (z \otimes x) \geq 0 \).
- Complete Positivity: The Choi matrix \( L \) is positive semi - definite.
2. **Research Objectives**:
- Look for specific types of linear matrix mappings so that their positivity and complete positivity are the same.
- Propose structural conditions under which the positivity and complete positivity can be equivalent.
3. **Main Results**:
- The author proposes a theorem (Theorem 1.1), pointing out that under certain structural conditions, the positivity and complete positivity of linear matrix mappings are the same.
- These conditions involve the selection of linearly independent matrices and their positional relationships (such as condition (C1)).
### Key Formulas
- Definition of Choi matrix \( L \):
\[
L = [L_{ij}] \in F^{nq \times nq}, \quad L_{ij} = L(E^{(q)}_{ij}) \in F^{n \times n}
\]
where \( E^{(q)}_{ij} \) is the standard basis element matrix, which is 1 at position \((i, j)\) and 0 elsewhere.
- Definition of linearized matrix \( L \):
\[
L: F^{q^2} \to F^{n^2}, \quad L(\text{vec}_q(V)) = \text{vec}_n(L(V)), \quad V \in F^{q \times q}
\]
- Positivity condition:
\[
(z \otimes x)^* L (z \otimes x) \geq 0 \quad \text{for all} \quad x \in F^n \quad \text{and} \quad z \in F^q
\]
- Condition (C1):
\[
\text{For each} \ k, \ \text{there is} \ j_l \neq j_k \ \text{for each} \ l \neq k \ \text{or} \ i_l \neq i_k \ \text{for each} \ l \neq k
\]
### Summary
The main contribution of this paper lies in linking the positivity and complete positivity of linear matrix mappings through structural conditions and proving that in some specific cases, these two properties are the same. This is of great significance for understanding and applying the theory and practical problems of linear matrix mappings in multiple fields.