A Game Theory Approach to Robust Discrete-Time H-Estimation

Y. Theodor, Uri Shaked

Research output: Contribution to journalArticlepeer-review


The deterministic robust estimation problem is formulated in the discrete-time case as a zero-sum two person difference game, in which a statistician plays against nature. The statistician tries to estimate a linear combination of the states of an uncertain linear system, which is driven by nature. Saddle-point strategies for the statistician in the above game are difficult to find. A suboptimal strategy is, therefore, considered which guarantees an Hperformance bound on the estimation error. Different patterns of the information that is available to the statistician are treated, and the corresponding fixed-point, fixed-lag. and fixed-interval smoothing strategies are derived.

Original languageEnglish
Pages (from-to)1486-1495
Number of pages10
JournalIEEE Transactions on Signal Processing
Issue number6
StatePublished - Jun 1994


Dive into the research topics of 'A Game Theory Approach to Robust Discrete-Time H-Estimation'. Together they form a unique fingerprint.

Cite this