Common-solvability conjecture for systems of linear matrix inequalities

At least 12 years old · documented by

Let A1,…,AkA_1,\ldots,A_k and B1,…,BkB_1,\ldots,B_k be matrices of compatible sizes, and let XX range over Hermitian matrices. Consider the kk linear matrix inequalities

B1XB1∗≽A1,…,BkXBk∗≽Ak.B_1XB_1^{*}\succcurlyeq A_1,\quad \ldots,\quad B_kXB_k^{*}\succcurlyeq A_k.

Common-solvability conjecture. The kk LMIs

B1XB1∗≽A1,…,BkXBk∗≽AkB_1XB_1^{*}\succcurlyeq A_1,\quad \ldots,\quad B_kXB_k^{*}\succcurlyeq A_k

have a common Hermitian solution if and only if each of the kk LMIs has a Hermitian solution individually.

The preceding two-LMI result establishes this equivalence for k=2k=2. The conjecture asks whether separate solvability always implies simultaneous solvability for an arbitrary finite family of LMIs of this form.

References

Primary source

Yongge Tian, “Analytical solutions to some optimization problems on ranks and inertias of matrix-valued functions subject to linear matrix inequalities”, arXiv:1301.0986 (2013).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.