Auto-Determining Mainlobe Width for Beampattern Synthesis via Relaxation Optimization

Thispaper presents a novelarray pattern synthesis algorithm based on relaxation optimization. We consider the problem of how to auto-determine the width of mainlobe region in beampattern synthesis. By introducing a slack region and a slack vector, the beampattern synthesis problem can be transformed as a relaxation optimization problem. The relaxation optimization problem is non-convex with an objective of sparsifying the slack vector. The sequential convex optimization procedure is then applied to solve the relaxation optimization problem and find its sparse solution. Representative simulations are provided to demonstrate the effectiveness of the proposed method in beampattern synthesis.