A possible method of classifying all density matrices for a N-state quantum system into different types, using the sameness of the characteristic polynomial of density matrices belonging to a given type, is discussed. The method employs the SU(N) generator expansion of density matrices. It is demonstrated that density matrices belonging to a given type, although different, share many common properties including the same eigenvalues, the same shape of the region in the parametric space in which lie the allowed parameter values, and the same global averages of the pertinent observables. Furthermore, we discuss the time evolution of a given type of density matrices and show that, in the appropriately defined time-dependent comoving matrix basis, a density matrix of a given type belongs, during entire unitary evolution, to the same type. In this way one is able to group and organize the multitude of different density matrices with certain analogous properties, into a smaller number of types, thus systematizing and cataloguing different mixed and pure states of a N-state system. The general discussion is illustrated throughout with the special case of a N=4-state quantum system. Two- and three-dimensional cross sections of the space of generalized Bloch vectors are determined by the Monte Carlo sampling method for several types of density matrices, providing some insight into the intricate and complex geometric structure of the space of density matrices.
The properties and structure of the eight-dimensional parameter space of three-state density matrices are examined using the SU(3) generator expansion. Three-dimensional cross sections of the space of generalized Bloch vectors are determined by the Monte Carlo sampling method and an appropriate classification of 56 different three-parameter density matrices into ten types, according to the shape of the corresponding region of allowed values of three relevant parameters, is introduced. Positions of the representative points corresponding to the pure states are determined for all ten types.
Determining whether quantum states can be interconverted by a prescribed set of transformations is a fundamental problem in quantum information. Here we develop a general classification framework based on conserved operators (COs), i.e., not necessarily hermitian matrices having the state of interest as an eigenstate. We first show that COs belonging to the algebra generating the allowed transformations can be obtained from the kernel of a quantum covariance matrix. Their number determines the dimension of the accessible orbit, while their algebraic structure and eigenvalues provide further constraints on state convertibility. We establish general bounds on the number of COs and prove that highest-weight states, like separable states in the case of
n qudits, attain the maximum. We also show how these operators generate parent Hamiltonians having the characterized state as an exact ground state. We then also consider positive zero-eigenvalue COs, which encode forbidden directions of reduced density matrices and incorporate rank-based constraints into the same framework. Applying this approach to multiqubit SLOCC, we obtain complementary classifications from algebraic and rank COs. While for
n≤3 qubits both lead to the usual classes, for
n=4 qubits they distinguish 17 and 20 types separately, and 29 when combined. The combined classification yields 287 fixed types and
4 continuous families for
n=5 qubits, and 4926 fixed types and 147 continuous families for
n=6 qubits, providing, to our knowledge, the most refined exhaustive classification of SLOCC-convertibility classes based on such data.
This investigation continues a program aiming at obtaining effective quantum models to describe measurement statistics. In [Stark, arXiv:1209.5737 (2012)], we have described how the Gram matrix associated to the prepared states and the performed POVM elements can be estimated via convex relaxation of a rank minimization problem. This Gram matrix determines the density matrices and the POVM elements uniquely up to simultaneous rotations (with respect to the Hilbert-Schmidt inner product) in the vector space of Hermitian matrices. However, when the description of the experiment needs to be connected to textbook quantum mechanics, explicit expressions for the states and the POVM elements in terms of positive semidefinite matrices are required. In this paper, we describe a heuristic algorithm that takes the state-meaurement Gram matrix as input and searches explicit realizations of this Gram matrix in terms of quantum mechanically valid density matrices and POVM elements.