Optimization of a Constrained-Feed Rotman Lens Beamformer
Abstract
A previous mathematical analysis has demonstrated the utility of a partially overlapped, constrained-feed network for time-delay control of large linear arrays (R.J. Mailloux, IEEE Trans. 49, February 2001, pp.280-291). In particular, this novel method allows for approximately -30 dB sidelobe suppression over a 20% bandwidth. An array with time-delayed contiguous subarrays with the same separation would have quantization lobes at the -10 dB level; thus, this technique appears to offer significant advantages. We recently developed an experiment to demonstrate this concept. We collected data for the broadside case (array phase shifters set to zero) from -45 degrees to 45 degrees in 0.25 degrees increments and from 9.0 to 11.0 GHz in 0.05 GHz increments (center frequency of 10 GHz). We allowed weighting of both the Rotman lens outputs (constituent beams) and the Butler matrix outputs (subarray patterns). We used a genetic algorithm to optimize these complex weightings. We realized that our data set didn't represent the best measure of system performance, since there is no beam squinting at broadside; therefore, we performed a limited field of view (LFOV) test. Using this LFOV test, we were able to demonstrate at least -28 dB sidelobes over an angular field of view corresponding to a 20% bandwidth.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 20, 2005
- Accession Number
- ADA486592
Entities
People
- James P. Kenney
- Michelle H. Champion
- Robert J. Mailloux
- Scott G. Santarelli
Organizations
- Air Force Research Laboratory