Optimization of array tapers subject to transmit constraints

Array tapers are traditionally used only on receive, as classical low-sidelobe tapers would result in a significant reduction in transmitted power and most transmit arrays lack element-level amplitude control. As modern arrays move towards solid-state transmitters distributed to each element there is a desire to more fully harness this flexibility. It has been previously shown that the transmit power loss can be reduced by designing tapers that upper-bound the coefficients while allowing a small number of inner array-factor sidelobes to remain unconstrained. That work is extended here to remove structural limitations of the prior algorithm, admit arbitrary combinations of peak and mean-square constraints on the weights and the array factor, allow controlled “overdriving” of some elements, and allow lower-bounding of the array weights to limit the dynamic range of the taper. These extensions are enabled though the use of second-order cone programming, a popular form of constrained convex optimization.