The least extremal solution of the operator equation \(AXB=C\) (Q2638490)

From MaRDI portal
scientific article
Language Label Description Also known as
English
The least extremal solution of the operator equation \(AXB=C\)
scientific article

    Statements

    The least extremal solution of the operator equation \(AXB=C\) (English)
    0 references
    0 references
    1990
    0 references
    Throughout his theses \({\mathcal B}\) denotes a Banach space, \({\mathcal H}\) is a Hilbert space. Let L(\({\mathcal B}_ 1,{\mathcal B}_ 2)\) and \({\mathcal C}({\mathcal B}_ 1,{\mathcal B}_ 2)\) be the sets of all bounded linear operators and space of all densely defined closed linear operators respectively. In 1956, using Frobenius matrix norm, R. Penrose proved that the least extremal solution of the matrix equation \(AXB=C\) is a matrix \(A^+CB^+\), where \(A^+\) is the Moore-Penrose inverse. In this theses, the author generalized Penrose's result to the operators in infinite dimensional complex Hilbert spaces. To do this the author considered an operator \(A^{\square}B\) defined by \(A^{\square}BX=AXB\), both when A, B are bounded linear operators and when A, B are arbitrary closed linear operators. In the former case, \(A^{\square}B\) has a generalized inverse \((A^{\square}B)^+\) and \((A^{\square}B)^+=A^{+\square}B^+\), in the later case \(A^{\square}B\) has a topological generalized inverse. Using the above results, the author proved that 1) If \(A\in L({\mathcal H}_ 3,{\mathcal H}_ 4)\), \(B\in L({\mathcal H}_ 1,{\mathcal H}_ 2)\) then the least extremal solution (with respect to the Hilbert- Schmidt norm) of the equation \(AXB=C\) is the operator \(\hat X=A^+CB^+\), where \(C\in C_ 2({\mathcal H}_ 1,{\mathcal H}_ 4)\), \(X\in C_ 2({\mathcal H}_ 2,{\mathcal H}_ 3)\) and \(C_ 2\) is the class of all Hilbert-Schmidt oprators. 2) If \(A\in {\mathcal C}({\mathcal H}_ 3,{\mathcal H}_ 4)\), \(B\in {\mathcal C}({\mathcal H}_ 1,{\mathcal H}_ 2)\) then the least extremal solution (with respect to 2-norm) of the equation \(AXB=C\) is \(\hat X=A^+CB^+\), where C and X vary in some specified spaces.
    0 references
    0 references
    Frobenius matrix norm
    0 references
    matrix equation
    0 references
    Moore-Penrose inverse
    0 references
    Hilbert- Schmidt norm
    0 references
    Hilbert-Schmidt oprators
    0 references
    least extremal solution
    0 references
    0 references