1: \begin{abstract}

2: 

3: A conventional monopulse radar system uses three beams; sum beam, elevation difference beam and azimuth difference beam,

4: which require different layers of weights to synthesize each beam independently.

5: Since the multi-layer structure increases hardware complexity, many simplified structures based on a single layer of weights

6: have been suggested.

7: In this work, we introduce a new technique for finding disjoint and fully covering sets of weight vectors, each of which

8: constitutes a sparse subarray, forming a single beam.

9: Our algorithm decomposes the original non-convex optimization problem for finding disjoint weight vectors

10: into a sequence of convex problems.

11: We demonstrate the convergence of the algorithm

12: and show that the interleaved array structure is able to meet difficult beam constraints.

13: %in comparison with other simplified weight structures.

14: 

15: \end{abstract}

16: