Incorporating Light Beam Search in a Vector Normal Boundary Intersection Method for Linear Antenna Array Optimization

An high directivity and low side lobe level antenna array is essential in wireless communication systems. Currently, the array design is often based on the experience of the designer. To combine the preferences of a decision maker into the evolution process of a multi-objective linear antenna array design problem, the light beam search method is proposed to be incorporated into a vector normal boundary intersection method to find an arbitrary fraction of the Pareto front. To adapt different preferences, a new transformed projection mechanism is introduced. To enhance the diversity of the algorithm, a dynamic transitional sub-domain strategy is proposed and used. Typical mathematical test functions and a ten-element linear antenna array optimization problem are used to testify the effectiveness of the proposed method.