EAGER: Advanced Development of Genetic Programming for Novel Active Metamaterials and Devices in Terahertz (THz) Regime

University Of Hawaii, Honolulu

Investigators

Abstract

The terahertz frequency band spans from 0.1 to 10 terahertz which is much (hundreds of times) higher than the radio frequency band commonly used for wireless communications. Since higher frequencies can carry larger data, the terahertz band ensures much higher data rate for wireless communications which is very much desired in modern times. But it is very difficult to design and build devices working in the terahertz regime. This is due to the fact that most natural materials do not respond to terahertz radiations. Therefore, artificial materials or metamaterials have become mainstream in the development of terahertz devices. Current design of metamaterials and devices is mostly based on traditional methods and heavily depends on expert knowledge in this area. This EAGER project proposes to use genetic programming for the design and optimization of the active metamaterials and devices in the terahertz band. Genetic programming is an advanced evolutionary optimization method in which the solution is represented as a computer program and has produced human-competitive, novel designs in many fields of engineering. It is more advanced than the well-known genetic algorithm which is often used to optimize parameters in a given design so as to meet set specifications. Genetic programming does not need pre-defined topology of the design. Instead, genetic programming can figure out the optimal topology as well as other specifications. The result of this EAGER project includes novel algorithms of genetic programming and the much needed designs of active metamaterials and devices in the terahertz frequency band. The objective of the proposed work is to fully develop Genetic Programming (GP) algorithm and optimization method for designing novel active metamaterials for practical device and components implementation in the Terahertz (THz) band. Technologies in this band have significant applications in developing devices for high data rate wireless communications as well as for medical imaging, explosives detection, security screening, sensors, and so on. Development of active metamaterials in the terahertz band is critically important for the evolution of these technologies and realization of their much anticipated benefits. The multitude of possible designs of active metamaterials for variety of device technologies leads to significant process complexities that are infeasible to explore manually. Genetic Programming, on the other hand, is an advanced evolutionary optimization method and has produced human-competitive, novel designs in many fields of engineering. Earlier GP work of the research group was focused on the design of challenging broadband Artificial Magnetic Conductor (AMC) metamaterial ground plane at lower frequency (a few hundreds of MHz). Through true 3D patterning several successful broadband designs were achieved in the ?no-man?s? frequency band. Matching genetic programming with active metamaterials development, therefore, presents a commendable approach, especially in the terahertz band, where natural materials are of limited application. Specific tasks of the proposed work include: 1) Using GP to develop active metamaterials and terahertz (THz) devices; 2) Improving the computational efficiency of GP through parallelization and implementation of more efficient (gradient based) optimization algorithms; and 3) Develop examples of optimized terahertz devices including detectors, modulators, sensors, and antenna arrays; and compare performance results with available designs based on human expertise. This work will also result in significant advances in Genetic Programming development including topology generation, tunable material integration, and improved computational efficiency that can tackle the high complexities involved in active aspects of the metamaterials and devices developments.