Miniaturized biocompatible microstrip antenna using genetic algorithm

Biocompatible antennas are an area of recent research that can facilitate remote communication with medical implants. This paper shows several possible designs of a "waffle-type" antenna created using genetic algorithms (GAs) that are of a size potentially suitable for cardiac pacemakers in the 402-405 MHz MICS band. In addition, methods to constrain the simulation and speed up the convergence of the GA for this type of antenna are explored. Even simple constraints such as fixing the feed and ground locations and encouraging the antenna to grow preferentially in the horizontal direction to take advantage of the longer physical dimensions of a pacemaker in that direction can appreciably improve convergence speed. One exception to this was constraining the patches to be connected together rather than distributed randomly, which caused the system to converge slower rather than faster. Also notable was that when a larger physical size was allowed, the system also converged more quickly despite a sizeable increase in the number of unknowns (subpatches) in the model. Thus, this paper provides a smaller, better matched microstrip antenna for biotelemetry and a choice of GA constraints for designing it.