Convex Optimization for Low-Complexity Spatial Antenna Arrays

Abstract

Conventional smart antennas create flexible beam patterns using weights that have both magnitude and phase. This requires expensive hardware in the form of individual receivers and transmitters, high-speed D/A and A/D converters and capable DSP or FPGA processors. This paper looks at the use of low-complexity spatial antenna arrays that can create reasonably complex antenna patterns using a phased array approach. Convex optimization is applied to solve the highly non-linear optimization problem. Square array geometries were studied in depth by applying various penalty functions. We found that convex optimization is a novel and effective way to compute the complex antenna weights, and that this low-complexity approach is an interesting alternative to more expensive smart antennas