Title: Some recent results in decentralized H2 optimal control subject
to delay

Abstract: In the past decade, both the state and output-feedback
problems have been solved for H2 optimal control subject to delay.
All of these solutions assume (1) a given quadratically invariant (QI)
communication pattern and (2) perfect, but fixed, delayed
communication between individual plants.  Our recent work has focused
on relaxing all of these assumptions, with the ultimate goal of
closing the gap between communication (both coding and network
design), and control.  In particular we have developed heuristics for
a principled suboptimal design when the communication pattern is not
QI (inspired by potential smart-grid applications); exploited a dual
problem formulation to not only further analyze the output-feedback
problem, but also to inform communication network augmentation; and
finally, we have extended dynamic programming solutions to the
state-feedback problem such that it can accommodate time-varying (but
bounded) delays.