“Diversity-Integration Trade-offs in MIMO Radars”
Abstract
Multiple-Input Multiple-Output (MIMO) Radars with widely spaced antennae at both the transmitter and the receiver may achieve substantial performance improvements over conventional systems by
1.Generating
as many independent and identically distributed
replicas of the target echoes as possible, thus
fully exploiting the diversity granted by the
different aspect angles;
2.
Integrating energy along fewer paths, i.e. giving up
the transmit diversity in favor of stronger signal
on each surviving path;
3.Compromising
between [a] and [b].
In the above framework, the present talk is aimed at shedding some light on the operation of MIMO radars. The topics covered in the talk can be summarized as follows:
1.Assuming
that each transmit antenna is assigned an
N−dimensional code-word, a general model for the
received signal is presented;
2.At
the receiver design stage, the Generalized
Likelihood Ratio Test (GLRT) is derived for
arbitrary transmitted signals, showing that it
exhibits a canonical structure, namely always
consists of a projector followed by an energy
detector;
3.The
availability of closed-form formulas for the
false-alarm and detection probability under Gaussian
target scattering and arbitrary noise
time-correlation allows showing that suitable design
of the code-book employed at the transmitter allows
trading diversity for integration. Since no
uniformly optimum (i.e., for any
signal-to-disturbance ratio) strategy exists,
space-time coding turns out to be a precious degree
of freedom at the system design stage.
4.Some
considerations on the case of non-Gaussian
scattering will be presented;
5.A
number of further developments and hints for future
research will be given.
|