Modified and Hybrid Cuckoo Search Algorithms via Weighted–Sum Multiobjective Optimization for Symmetric Linear Array Geometry Synthesis

Abstract: This study proposes the modified and hybrid cuckoo search algorithms deploying the weighted–sum multiobjective optimization approach in synthesizing symmetric linear array geometry with high directivity, low average side lobe level (SLL), a small half–power beamwidth (HPBW), and/or significant predefined nulls mitigation. The weighted–sum approach optimizes three objective functions simultaneously until the maximum number of iteration achieved. Precisely, the modified cuckoo search (MCS) algorithm is introduced through the integration with the Roulette wheel selection operator, the adaptive inertia weight controlling the positions (solutions) exploration, and the dynamic discovery rate of solutions. Besides, there are also the proposals of hybrid MCS with two popular evolutionary algorithms, which are the particle swarm optimization (PSO) known as MCSPSO and the genetic algorithm (GA) referred as MCSGA. All the modified and hybrid cuckoo search–based multiobjective algorithms go through the weighted–sum approach to generate three optimal decision variables, which are array element excitation locations, amplitudes, and phases, respectively. The optimal solutions obtained through various MATLAB simulations are then compared against corresponding counterparts.

Keywords: Modified Cuckoo Search Algorithm, Particle Swarm Optimization, Genetic Algorithm, Weighted–Sum Multiobjective Optimization, Side Lobe Level Suppression, Half–Power Beamwidth, and Nulls Control.