IEEE SIGNAL PROCESSING LETTERS, VOL. 11, NO. 11, NOVEMBER 2004 875
Cramer–Rao Lower Bound for Constrained
Complex Parameters
Aditya K. Jagannatham, Member, IEEE, and Bhaskar D. Rao, Fellow, IEEE
Abstract—An expression for the Cramer–Rao lower bound
(CRB) on the covariance of unbiased estimators of a constrained
complex parameter vector is derived. The application and useful-
ness of the result is demonstrated through its use in the context of
a semiblind channel estimation problem.
Index Terms—Channel estimation, constrained parameters,
Cramer–Rao bound (CRB), MIMO, semiblind.
I. INTRODUCTION
T
HE CRAMER–RAO lower bound (CRB) serves as an
important tool in the performance evaluation of estima-
tors which arise frequently in the fields of communications
and signal processing. Most problems involving the CRB are
formulated in terms of unconstrained real parameters [1]. Two
useful developments of the CRB theory have been presented in
later research. The first being a CRB formulation for uncon-
strained complex parameters given in [2]. This treatment has
valuable applications in studying the base-band performance
of modern communication systems where the problem of
estimating complex parameters arises frequently. A second
result is the development of the CRB theory for constrained
real parameters [3]–[5]. However, in applications such as
semiblind channel estimation one is faced with the estimation
of constrained complex parameters. Though one can reduce the
problem to that of estimating constrained real parameters by
considering the real and imaginary components of the complex
parameter vector, the complicated resulting expressions result
in loss of insight. Using the calculus of complex derivatives as
is often done in signal processing applications, considerable
insight and simplicity can be achieved by working with the
complex vector parameter as a single entity [1], [6], [7]. We
thus present an extension of the result in [3]–[5] inspired by the
theory in [2] for the case of constrained complex parameters.
To conclude, we illustrate its usefulness by an example of a
semiblind channel estimation problem.
II. CRB FOR COMPLEX PARAMETERS WITH CONSTRAINTS
Consider the complex parameter vector . Let
such that the real and imaginary parameter vectors
Manuscript received February 2, 2004; revised April 7, 2004. This work was
supported by CoRe Research Grant Cor00-10074. The associate editor coordi-
nating the review of this manuscript and approving it for publication was Prof.
Steven M. Kay
The authors are with the Center for Wireless Communications (CWC),
University of California at San Diego, La Jolla CA 92093 USA (e-mail:
jak@ucsd.edu).
Digital Object Identifier 10.1109/LSP.2004.836948
and . Assume that the likelihood func-
tion of the (possibly complex) observation vector pa-
rameterized by is . Let be given as
, where are unbiased estimators of ,
respectively. In the foregoing analysis, we define the gradient
of a scalar function as a row vector:
(1)
Let be defined as in [2] by
(2)
Suppose now that the complex constraints on are given as
(3)
i.e., . We then construct an extended constraint set
(of possibly redundant constraints) as
(4)
An important observation from (4) above is that symmetric
complex constraints on these parameters are treated as disjoint.
For instance, given the orthogonality of complex parameter
vectors , i.e., , the symmetric constraint
is to be treated as an additional complex con-
straint and hence . The extension of
the constraints is akin to the extension of the parameter set
from to called for when dealing with complex
parameters, and the need will become evident from the proof
of lemma (1). Reparameterizing in
terms of , let the set of parameter constraints for be given
by . Employing notation
defined in [3] and borrowing the notion of a complex derivative
from [1], [6], we define as
(5)
It then follows from the properties of the complex derivative [6]
that
(6)
1070-9908/04$20.00 © 2004 IEEE