Search Network

Global Optimisers for World Winde Antenna Demands

30 November 2011
print version share on facebook share on linkedin

Executive summary

The realisation of enhanced bandwidth on a printed monopole antenna was achieved with a Multi-objective Evolutionary Algorithm. The antenna radiating element and ground-plane features, both based on Bézier spline outlines, were simultaneously optimised for best radiating performance.

Challenge overview

Evolutionary optimisation methods such as Genetic Algorithms (GAs) and Particle Swarm Optimisation (PSO) have been successful applied to electromagnetic design. Such techniques attracted attention due to their ease of fit to a variety of problems. The algorithms initialise with random solutions from a large 'search space' and then proceed to evolve haphazardly towards an optimum using a fitness function is used to assess the merits of possible outcomes results. The computational time for trialling fitness of a single solution varies from a few seconds to tens of minutes. Accordingly, the optimisation can depend on hundreds or thousands of solver solutions which can result in days or weeks of experimental time. A more efficient approach can realise faster solutions.

Implementation of the initiative

The Dublin Institute of Technology's Antenna & High Frequency Research Centre uses an efficient global optimisation algorithm which is designed to solve multi-objective problems like those found in antenna design. The algorithm is suited for scenarios where computational power is limited or is restricted to a single commercial license and thus, parallelisation is not an option.The geometry of the antenna is constructed using quadratic Bézier spline curves. The use of the spline shapes for both radiator and ground-plane leads to a less constrained geometry, with a greater freedom for the optimisation and can provide resultant forms with less corners/edges and significantly improved antenna performance.DecaWave Ltd and Taoglas Ltd required miniaturised antennas with ultra wideband performance for their respective applications. Antenna performances function as a result of their installed environments and considerations took account of their customers' requirements, with discussion between offices across many time-zones.The final viable solutions were pre-defined by limits in costs, manufacturing techniques, ergonomics and basic antenna functionality. Such competing parameters were managed more efficiently by the new optimisation process, which had been refined on a PhD research programme. The principle challenge stemmed from trying to ensure that the predicted antenna performance would integrate with the yet to be finalised electronic circuitry and would avoid subsequent requirements re-design before deployment.

Results and achievements

The optimised antenna shapes produced best-in-class performance over a very wide band of frequencies (worldwide Ultra Wideband standards). One application involved real-time positioning systems where the antenna, which is device integrated, was optimised for best pulse fidelity, realising maximum positioning accuracy. Other applications for the patented antenna shaping method were the wireless communications sector, including an antenna for 4G/LTE, where the technique applied provided an antenna solution covering worldwide multi jurisdictional frequency bands.This work has resulted in licence agreements with DecaWave, Taoglas and others. Given that customer requirements are typically proprietary, a significant amount of engineering intervention is necessary in its use. However, the outcomes exceed competitor performances to increase reliability and reduce costs.Antenna showing surface currents at 14 GHz and spline outline with control points

Lessons learned and replicability

The development of a powerful design optimisation tool to balance competing parameters in antenna design has underpinned emerging radio technologies. The combination of a spline based geometry and the multi-objective evolutionary global optimisation algorithm are very powerful tools to optimise UWB antennas, especially if computational resources and time are limited.

Attached Documents

PDF icon Irish Mathematics and Industry (Irish_Mathematics_and_Industry.pdf | 1.42 MB)