Abstract
A new ramped convergence particle swarm optimization (RCPSO) algorithm to design radio frequency identification (RFID) reader antenna for gain optimization is presented in this paper. The algorithm breaks down the optimization problem by considering only a subset of dimensions at a time, hence overcoming curse of dimensionality. It uses a multi-start approach that alleviates premature convergence. It applies heterogeneous boundary conditions to errant particles. In this paper, the effectiveness of using RCPSO for RFID reader antenna design will be investigated by a number of iterations required to achieve optimum solution along with the quality of solution. RCPSO is built specifically for antenna design. For the microstrip antenna, a significant bandwidth improvement is achieved and the overall antenna dimensions are kept practically the same. The circular gain improves by a significant 2 dBiC over super high frequency (SHF).