Bilinear operator multipliers into the trace class

TL;DR

This paper characterizes bilinear maps between spaces of Hilbert-Schmidt and trace class operators using modular properties and von Neumann algebra tensor products, extending known results to a bilinear setting.

Contribution

It introduces a new framework for understanding completely bounded bilinear maps into trace class operators, generalizing classical linear results to the bilinear case.

Findings

Characterization of completely bounded module maps via von Neumann algebra tensor products.

Weak factorization property for maps into trace class operators when $M_2$ is injective.

New proof of Sinclair-Smith theorem using Hilbert $C^*$-modules.

Abstract

Given Hilbert spaces $H_1,H_2,H_3$, we consider bilinear maps defined on the cartesian product $S^2(H_2,H_3)\times S^2(H_1,H_2)$ of spaces of Hilbert-Schmidt operators and valued in either the space $B(H_1,H_3)$ of bounded operators, or in the space $S^1(H_1,H_3)$ of trace class operators. We introduce modular properties of such maps with respect to the commutants of von Neumann algebras $M_i\subset B(H_i)$, $i=1,2,3$, as well as an appropriate notion of complete boundedness for such maps. We characterize completely bounded module maps $u\colon S^2(H_2,H_3)\times S^2(H_1,H_2)\to B(H_1,H_3)$ by the membership of a natural symbol of $u$ to the von Neumann algebra tensor product $M_1\overline{\otimes} M_2^{op}\overline{\otimes} M_3$. In the case when $M_2$ is injective, we characterize completely bounded module maps $u\colon S^2(H_2,H_3)\times S^2(H_1,H_2)\to S^1(H_1,H_3)$ by a weak factorization property, which extends to the bilinear setting a famous description of bimodule linear mappings going back to Haagerup, Effros-Kishimoto, Smith and Blecher-Smith. We make crucial use of a theorem of Sinclair-Smith on completely bounded bilinear maps valued in an injective von Neumann algebra, and provide a new proof of it, based on Hilbert $C^*$-modules.