Recursive Cramer Rao Lower Bound for Random Parameters
In many cases an estimator is needed to estimate a certain quanmtity from an obser- vation.
The estimators can take different forms and so a method of comparison or a method of
assessment will prove to be useful. The Cramer Rao Lower Bound (CRLB) is one such
method. The CRLB lower bounds the variance of an estimator; however, the CRLB assumes
the parameter to be estimated to be deterministic. The Posterior Cramer Rao Lower Bound
(PCRLB) bounds the variance of an estimate for stochastic parameters. In this paper the
PCRLB is thoroughly discussed and is also applied to the Autoregressive Moving Average