Element failure in antenna arrays leads to severe degradation of the radiation pattern. It is significantly more complex in circular antenna arrays (CAA) considering its geometry and elements spatial distribution along the circumference of the circle. Several design parameters like the coefficient of current amplitude excitation, phase, and arc length are determined appropriately to obtain the desired parameters involved in the pattern restoration process. It is possible to rely on alternate desired parameters to restore the patterns without disturbing the actual design parameters involved in the optimization process. In this paper, such an attempt to optimize CAA using non-uniform amplitude distribution is considered. After enforcing multiple-element failure, the resultant defective pattern is restored by determining the non-uniform spatial distribution of four consecutive elements preserving the corresponding non-uniform amplitude distribution. Hence, it is justifying mentioning that the position-only technique is used to restore the pattern in terms of optimized SLL and BW. The spatial distribution of optimized CAA is determined using the latest modified Social Group Optimization Algorithm (MSGOA) and compared the results with the well-known Flower Pollination Algorithm (FPA). The MATLAB® is used to carry out the simulations, and results are analyzed.