On array failure mitigation using genetic algorithms and a priori joint optimization

Abstract

Antenna arrays are widely used in communications systems, especially in base-station installations, mainly for improvement of system capacity and spectral efficiency. The possibility of an antenna-array element failure is examined in this paper. If an element fails during operation, a significant degradation in the array's performance is observed, even after the redistribution of the excitation coefficients of the remaining elements. In recent research in the field, the redistribution of the elements' excitation coefficients is performed after the array failure, and the possibility of failure is not taken into account during the design of the array. In this paper, the possibility of a one-element failure is predicted and properly treated in early design stages. If the array is initially designed taking into account the possibility of a one-element failure, the elements' coefficients are redistributed more efficiently and the failure mitigation is superior. In order to demonstrate the applicability of the proposed technique, two eight-element, switched-beam circular arrays, covering the horizontal plane, were designed using the method of genetic algorithms (GAs). The array elements were considered to be identical dipoles, placed vertically to the x-y plane. The beamwidth of the main-beam lobe was 45/spl deg/ (2/spl pi//N where N is the number of the array elements), the while maximum relative side lobe level (RSLL) was intended. Numerical results showed excellent agreement with the design goals.