Convex solutions of the direct data domain least squares approach in smart antennas

Convex solution frameworks for the direct data domain least squares (D3LS) weight coefficients in smart antenna arrays are presented. The D3LS mathematical relationships are decomposed into real and imaginary components and the corresponding relationships are formulated in linear programming and general convex optimization frameworks to achieve optimal weight coefficients within ε-accuracy. Furthermore, the concept of hypothetical jammer constraints has been introduced to synthesize more favourable beam patterns. The simulation results demonstrate that the convex solution frameworks are computationally very efficient and suitable for additional antenna system constraints, and perform well in the case of ill-conditioned matrices.