11.07.2015 Views

Principles of Modern Radar - Volume 2 1891121537

Principles of Modern Radar - Volume 2 1891121537

Principles of Modern Radar - Volume 2 1891121537

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

3.4 Optimum MIMO Design for Target Identification 99where⎡E { ˜H c ′ ˜H} −1 1c = ⎢P c ⎣⎤d 1 0 ··· 00 d 2 . ..⎥⎦0 ··· d N(3.37)andd i= (N + i − 1)−1(3.38)It is readily verified that the solution to (3.36) yielding the maximum eigenvalue is given by⎡ ⎤10s = ⎢⎣⎥(3.39). ⎦0Thus the optimum pulse shape for detecting a point target is itself an impulse. This should beimmediately obvious since it is the shape that excites the range bin only with the target andzeros out all other range bin returns that contain competing clutter.Of course, transmitting a short pulse (much less an impulse) is problematic in the realworld (e.g., creating extremely high peak power pulses) thus an approximation to a short pulsein the form <strong>of</strong> a spread spectrum waveform (e.g., LFM) is <strong>of</strong>ten employed [12]. This examplealso illuminates that in uniform random clutter nothing is gained by sophisticated pulse shapingfor a point target other than to maximize bandwidth (i.e., range resolution) [21]. The interestedreader is referred to [19] for further examples <strong>of</strong> optimizing other DOF (e.g., angle-Doppler)for the clutter mitigation problem.Up to this point we have been focused on judiciously choosing the transmit/receiveDOF to maximize SINR or SCR. In the next section we will extend this framework to thetarget identification problem.3.4 OPTIMUM MIMO DESIGN FOR TARGETIDENTIFICATIONConsider the problem <strong>of</strong> determining target type when two possibilities exist (the multitargetcase is addressed later in this section). This can be cast as a classical binary hypothesistesting problem [7]:(Target 1) H 1 : y 1 + n = H T1 s + n(Target 2) H 2 : y 2 + n = H T2 s + n(3.40)where H T1 , H T2 denote the target transfer matrices for targets 1 and 2, respectively. Forthe AGCN case, the well-known optimum receiver decision structure consists <strong>of</strong> a bank<strong>of</strong> matched filters, each tuned to a different target assumption, followed by comparator asshown in Figure 3-7 [7]. Note that (3.40) presupposes that either Target 1 or 2 is present, but

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!