The Optimization of Extreme Solutions for Antenna Designs

The attractiveness in low profile, structure simplicity, low production cost, lightweight, and RF circuitry conformability are all conformal within the design criteria of the microstrip antenna. However, despite these advantages, microstrip antenna does suffer from its inherent narrow bandwidth restriction (typically only a few percent), and which is not an easy undertaking task for designers.Furthermore, with the ever increasing demand for wireless communication/telecommunication services and the related applications for voice and data transmission, a desirably high data rate wireless transmission has already been attracting a great deal of attention. While the antenna is application orientated, the demands for the good performance in wide impedance bandwidth, reduced size, omnidirectional pattern, high gain (including uniform gain at individual frequencies), and polarization purity etc., still remains a Herculean task for designer to come with a design that can meet all these (some parts) of the criteria.Considering that the antenna design cycle that is usually painstakingly long even with the help of some accomplished computational tools and engineering skills, the required antenna performance that meets the design specifications must be further improved in order to maintain the requirement standard. The main design deficiency is largely because of (a) high-dimensional parametric variations, and (b) multiple design criteria.While genetic algorithm (GA) is a proven scheme for antenna optimization, a newly developed jumping genes paradigm, which can be regarded as a new form of evolutionary algorithm, is proposed as the principle design algorithm for the antenna design. This evolutionary scheme can address and produce solutions that can meet a range of antenna performance requirements. Its main feature of achievement is to allow the parameters/variables (genes) to jump from one location to other place within its chromosome structure, or even to its allied chromosomes. This scheme would alleviate the design difficulty of the current design methodology and that the antenna performance and as well as its architecture can be suitably optimized.The end-product of this project will form a designing platform which would easily become a commercially viable software tool to be readily adopted by a range of commercial/industrial design houses whenever the development of an antenna design is required.