![]() |
FIFTH
ANNUAL |
Signal Modulation in |
Daniel J. Rabideau MIT Lincoln Laboratory 244 Wood Street Lexington, MA 02173-9108 tel: (781) 981-2892 email: danr@ll.mit.edu Abstract Due to radar and jammer motion, the process by which an airborne jammer's energy is scattered off the terrain and into the receiver is continually changing. As a result, rapid weight updates have been proposed as a way to maintain interference suppression throughout a coherent processing interval. In fact, it has been suggested that the weights be re-computed on each pulse. In this presentation, we examine the effect of the "pulse-by-pulse" adaptive beamforming for hot clutter mitigation on other signals (e.g., targets or monostatic clutter) that may be present in the data. Two effects are examined in detail: random signal modulations induced by finite sample sizes and non-random signal modulations induced by the signals' coherence. In the first of these, overall signal localization in Doppler is degraded due to random fluctuations in the adaptive beampatterns (resulting from the use of estimated covariance matrices). Increased training set sizes or design constraints are effective at restoring localization performance. In the second effect, the coherence of the multipath interference gives rise to more predictable signal modulations. Unfortunately, these modulations cause the clutter ridge to expand within the angle-Doppler plane, requiring more adaptive degrees of freedom and more sophisticated STAP algorithms to reject. In this paper, we present an analytical model for both effects. We also compare and contrast these effects to the previously reported "clutter leakage" phenomena. Finally, we will discuss several techniques that might be employed to reduce the performance loss associated with these effects. |
Direct comments and questions to: webmaster@ll.mit.edu
© MIT Lincoln Laboratory. All rights reserved.