Nonlinear Dynamics of Adaptive Arrays for Improved Interference Suppression in Tactical Applications\
Rachel Goshorn · AIP conference proceedings · 2003
To date, the use of nonlinear dynamics of adaptive filter weights to provide improved spatial filtering has not been explored. Currently, interference suppression requires adaptation to a wide variety of signals, there may be multiple narrowband and broadband interfering signals from unknown directions. While adaptive arrays are used to control the spatial radiation pattern of antennas (typically by positioning nulls in the directions of interference sources), the methods are often thwarted by non‐stationary interference signals. In order to improve the current generation of adaptive arrays, one must now exploit the dynamic behavior of these arrays. Through the years, adaptive filters have been modeled to attain the classic optimal digital signal processing (DSP) Wiener filter performance, which has recently been shown to be a non‐optimal approach. In various applications, the weights of adaptive filters behave nonlinearly; they perform better than the optimal Wiener filter. Understanding this dynamic behavior allows for control of nonlinear effects to improve performance. Recent research shows optimal filter parameters (i.e. time constant, filter order) are significantly different than those predicted from classic adaptive filter theory. For example, classic analysis generally uses a small filter time constant to maximize performance. However, such small time constants suppress the weight dynamics, degrading the performance in important applications. Therefore, a new set of tools is required in the DSP community to understand the dynamics and improve performance. The nonlinear physics community has developed a number of powerful tools (e.g. phase models, bifurcation theory, order parameter equations, and delay‐embedding methods) to directly analyze the dynamics of nonlinear systems. These methods, along with digital signal processing analysis, will be applied to the adaptive array problem for performance improvements in spatial filtering.